Enhancements and defects fixed:

1. AI modified field indicator. AI Sparkle on Fields: AI-generated field indicators are now displayed consistently across both the platform UI and Workspace.
2. Complaint Case Intake Agent
   • Added the hyperlink to the complaint case number.
   • Prompt enhancements on location and products
   • Related Parties for Consumers.
3. Complaint Case Research Agent
   • Deleted default tasks after the creation of agent tasks.
   • Prompt enhancement on multi messages into one.
   • Search profile separation

Performance improvement.

Defect fixes and plugin upgrade issue fixed.

Defect fixes and plugin upgrade issue fixed.

Defect fixes and plugin upgrade issue fixed.

Defect fixes and plugin upgrade issue fixed.

Defect fixes.

Changed: AI skill execution is now performed off glide

Now Assist for code generation features now inherit skill-level ACLs configured in Now Assist Admin, ensuring restricted features are only available to authorized users.

Changed: Engagement tool for DEX agents to obtain employee approval before auto-triggering device actions. This change integrates with ZTS so will not be published until ZTS is GA.

Defect fixes and plugin upgrade issue fixed.

Fixed: Adoption of AI Gradient UI for integration builder.

Non Glide Cobalt Raven ACLs added.

New:

AI Framework that autonomously discovers schema, samples, and aggregates live ServiceNow tables, and presents a sourced answer without modifying any data.

Enhancements and defects fixed:

1. AI modified field indicator. AI Sparkle on Fields: AI-generated field indicators are now displayed consistently across both the platform UI and Workspace.
2. Complaint Case Intake Agent
   • Added the hyperlink to the complaint case number.
   • Prompt enhancements on location and products
   • Related Parties for Consumers.
3. Complaint Case Research Agent
   • Deleted default tasks after the creation of agent tasks.
   • Prompt enhancement on multi messages into one.
   • Search profile separation

Performance improvement.

Fixed:

Updated the version for ServiceNow Docs Connector to fix dependency failures

Fixed: AI Agent records view was affected due to an AI Native ACL issue

Defect fixes.

Now Assist for code generation features now inherit skill-level ACLs configured in Now Assist Admin, ensuring restricted features are only available to authorized users.

Changed

Engagement tool for DEX agents to obtain employee approval before auto-triggering device actions. This change integrates with ZTS so will not be published until ZTS is GA.

Fixed: Adoption of AI Gradient UI for integration builder.

Non Glide Cobalt Raven ACLs added.

New:

The Security Exposure 360 agentic workflow brings AI-powered exposure analysis to USEM. Users can now ask questions in plain language and get answers grounded in their own ServiceNow data across all types of findings within USEM.

New:

• Business users create templates to manage process so that all tasks are standardized across the organization. The template AI agent helps in creating templates from their current existing documents.
• Users can upload images or diagrams which contain a process flow/flow diagram. The Template AI agent reads these documents and analyzes the template and template items required.
• The Template ai agent creates a draft template along with template items based on the document and provides draft template details which are verified by the user.

New:

The Security Exposure 360 agentic workflow brings AI-powered exposure analysis to USEM. Users can now ask questions in plain language and get answers grounded in their own ServiceNow data across all types of findings within USEM.

No new features with this release. Updated the application to support bundling with other USEM applications.

New:

The Security Exposure 360 agentic workflow brings AI-powered exposure analysis to USEM. Users can now ask questions in plain language and get answers grounded in their own ServiceNow data across all types of findings within USEM.

New

• Added a new "Waiting for requester" activity state to support workflows requiring user follow-up or additional input.
• Added telemetry instrumentation for the in-product agentic experience to enable activity usage and engagement tracking.
• Introduced support for configurable visibility of the "Restart/Cancel Activity" option in the overflow menu to avoid conflicts with playbook activities.
• Completed foundational integration exploration between Suggested Actions AI Agent capabilities and the in-product experience.

Changed

• Updated headers across CSP, modal, and inline experiences from "AI Workflows" to "AI Activity" for improved terminology consistency.
• Renamed "Input required" to "Input needed" for clearer user messaging.
• Updated activity action text from "Cancel" to "Stop Activity."
• Refined follow-up action labels and completed several small UI and copy improvements to enhance overall usability and consistency.

Fixed

• Resolved workflow handling for requester wait-state scenarios to ensure activities properly reflect and manage pending user responses.
• Addressed overflow menu conflicts between activity controls and playbook activity options.

Removed outdated terminology and legacy labels associated with prior AI workflow naming and activity actions.

Changed:: Updated dependency from AI Foundation to FSO Now Assist

New application

• Discover automation opportunities across customer service interaction channels chat (portal and messaging), voice, and email in addition to incidents and cases.
• Transcript-based analysis infers customer intent and resolution steps from conversational language, surfacing automation opportunities that aren't captured in structured fields.
• Channel-specific agent matching recommends chat, voice, or email agents based on each opportunity's characteristics.
• Channel-appropriate agent generation when no suitable existing agent is found.

Changed

• Mining scope expanded beyond cases and incidents to include the Interaction table.
• Matching pulls from a unified agent registry so chat and voice agents are evaluated alongside existing agents and tools, with no separate setup per channel.
• Clustering now incorporates interaction type, transcript content, short description, and referenced knowledge base articles when ranking opportunities by impact.
• Fixed
  • Resolved a defect causing opportunity ranking to display in reverse order.
  • Resolved an issue where asset counts did not match across the UI.

New -

• AIOps AI Specialist (limited availavilblity) - The AIOps AI Specialist AI is ServiceNows purpose-built AI agent for IT Operations, designed to autonomously monitor, correlate, triage, and resolve infrastructure and application events at enterprise scale.
• Onboarding of the AIOps AI Specialist, including setting up the scope for the AI Agent, testing it and confgiure its task
• Remediation agent that works alongside the AIOps AI Specialist
• Proactive grouping recommendation using AI

Mosaic migration done for couple of Skills

Product Release Content Generator

Renewal Insight Engine

New:

AI Agent that investigates the missing attribute brought in through Pattern based discovery, finding affected CIs and links errors to standard error codes.

Updated to support the latest version of the dependent apps.

Fixed: Missing xml field that were causing data quality tests to fail

In this version, the following enhancements have been made:

• Security and access control management have been enhanced with simplified installation and automatic configuration handling.
• Stockroom locations no longer appear for software requests.
• The Orchestrator now displays human-readable output in work notes.
• Document extraction now works seamlessly with improved field configuration.

New

Added Splunk, AWS CloudWatch, ThousandEyes, SolarWinds, and Prometheus as supported investigation backends for the Analyze alert impact agentic workflow, expanding the alert insights available.

• Updated the Analyze alert impact agentic workflow to route all alert investigation through a single SRE Investigate AI Agent, delivering consistent behavior across all supported backends and simplifying the path for adding new vendor agents.
• Migrated New Relic and Datadog integrations to Model Context Protocol (MCP) agents, standardizing on vendor-maintained tool surfaces so the investigation workflows stay current as vendors ship updates.

Deprecated

Deprecated the Dynatrace Analysis AI Agent, superseded by the Dynatrace MCP Server Agent.

Connections tab in the Service Exchange Center

Create, view, request, and offboard provider and consumer connections from a single location in the Service Exchange Center. Search and filter connections without navigating across multiple screens.

Improved consumer registration and onboarding

Onboard consumers faster with a guided, step-by-step registration experience. Upgraded consumers are automatically redirected to this experience to receive clearer progress indicators during onboarding, and actionable messaging for failure and delay scenarios, minimizing onboarding friction and support dependency.

Improved FDS capabilities

• Improve your connection experience, by syncing Knowledge Base articles between provider and consumer instances.
• Reduce data inconsistencies by maintaining sys IDs for CMDB data and dependent relationships through transform maps.
• Ensure CI functionality is preserved on the destination instance by choosing to automatically create CI dependency relationships when relationship data is received from the source.
• Improved compliance through restricted data sync from non production instances to production instances for CMDB tables.

Journal Field Framework enhancements

• Increase flexibility in journal data synchronization between provider and consumer instances by mapping multiple source fields to a single target journal field.
• Configure journal fields such of type journal_input fields alongside journal type, ensuring all journal entries are preserved during synchronization without requiring custom scripting

Group-based persona assignments for Remote Catalog

Assign Remote Catalog personas to user groups so existing group-based access management practices extend to Remote Catalog, reducing administrative effort by managing access at the group level instead of individual users.

New

Added a Reliability Domain Agent that analyzes historical, post-incident data to identify service level objective (SLO) and synthetic monitoring gaps across configuration items (CIs) and services, then generates reliability tasks to address them.

Initial Release

• New
• Changed: Workplace Utilization QnA Agent can now be used by users with the sn_wsd_spcmgmt.space_planner role
• Fixed: The AI Agent was not always able to block the location after submitting an emergency maintenance request
• Removed

• New
  • Manage Smart Assessment templates with versioning support. Create, publish, and delete template versions to support consistent assessment governance.
  • View the entity, risk, and control for each assessment directly in task and work queue lists, without opening individual records.
  • Access authority documents, agency mappings, and citations for additional AI regulatory frameworks in the AI Risk and Compliance content pack.
• Changed: Added automated impact assessment flow whenever an AI Dataset is created.
• Fixed
  • Localization & Performance issues are fixed.
  • Fixed issue w.r.t to retired controls. Retired controls are excluded from the control based widgets.
  • Fixed issue related to synchronization of AI asset life cycle tasks between AI Risk and Compliance and AICT workspace.
  • Fixed functional domain issue w.r.t indicators feature on AI asset record page.

This is a "marker app" with no new functionality added; its purpose is to automatically trigger installation of other apps.

New:

• Refine actions You can now refine generated summaries directly within the dashboard using built-in actions such as Elaborate and Shorten, giving you more control over how the summary is presented.
• Feedback on generated content Added thumbs up / thumbs down controls to the summary component, making it easy to share feedback on AI-generated summaries.
• Dashboard Summary for UI Builder Workspaces Dashboard Summary is now available out of the box in UI Builder Workspaces.

Fixed:

• Consistent cache behavior across summary triggers Both the Refresh and Generate Summary buttons now apply the cache consistently, avoiding redundant AI calls.
• Dashboard Summary time zone aligned with user preference The Dashboard Summary now correctly applies the user's configured time zone rather than the system time zone, ensuring consistency with the data shown in dashboard widgets.

• Improved keyboard navigation in explorations:
  • Navigate explorations fully via keyboard, including text and action buttons.
  • Move up/down buttons have been added under the drag-and-drop icon.
• The data visualization sparkle icon is no longer shown for data sources in a non-supported secure scope.

• Record desktop actions more accurately by using the new AI-powered recording mode when creating desktop actions.
• Save time on manual setup by letting AI automatically insert anchors and generate screen context for each captured screen and add desktop action description after recording.
• Switch to AI-assisted recording by selecting the new Record with AI (recommended) check box that replaces the previous capture modes in the Create desktop action modal.
• Make desktop actions more flexible by configuring parameters for on-screen task desktop actions.
• Pass dynamic values at runtime by mapping parameters in the Map parameters section in AI Agent Studio.
• Control data visibility and security by using the Shared and Mark As Sensitive fields on the Desktop action parameter form.
• Get a quick guidance on how to effectively use the recorder with the recorder tips modal.

New features for June Release:

• Improved error handling and internationalization (i18n) support
• Added streaming capabilities for real-time data processing
• Resolved critical defects and stability issues
• Extended support for additional widget property types

• AI for Document Designer A new AI-powered application that integrates with the Microsoft Word add-in, enabling users to generate and manage report content directly within Word documents using Now Assist.
• AI-powered Word content generation A custom AI skill that generates Microsoft Word report content through natural language prompts, allowing users to create document sections conversationally.
• Content insertion into Word documents Support for inserting AI-generated content (HTML and OOXML formats) directly into Word documents via sn-office-addin handlers.
• Multi-conversation document editing Ability to edit and refine document content across multiple conversations with the AI assistant, enabling iterative content development.
• Data model, ACLs, and roles Purpose-built data model with access controls and role definitions to secure AI-generated document content and manage user permissions.

Changed

AI agent accuracy and prompt enhancements Improved agent accuracy and refined prompts based on Architecture Review Board (ARB) feedback for higher quality content generation.

Fixed: AI agent icons updated in add-in Corrected the icons displayed for the AI agent within the Microsoft Word add-in.

• Addressed Brazil security directive
• Honoured document_match_count in the Uber RAG instead of default.
• Updated default chunking mode to STB(Small to big) instead of Fixed Size.

Changed

Deprecated searching and scraping functionality.

New

AIOps AI Specialist (L1 AI Worker)

AI-powered first responder that autonomously triages alerts, reduces noise, enriches context, and provides remediation recommendations across the alert lifecycle.

• Context-rich alert investigation Enhanced alert analysis using CMDB topology, incident history, and operational signals to provide deeper insights and faster decision-making.
• Knowledge-driven investigation Integrated knowledge sources, including KB articles into the investigation workflow to improve accuracy and deflect incidents
• Remediation suggestions from existing subflows AI-generated remediation recommendations leveraging existing alert rules and subflows to accelerate resolution and ensure consistency.
• SRE AI Specialist integration Integrates seamlessly with the SRE AI Specialist for deeper investigation, including observability data analysis and root cause hypotheses

Dedicated AIOps Manager Homepage

New homepage for AIOps Managers and Event Management Admins to onboard, configure, and manage AI Specialists from a centralized command center. _

_

New:

Proactive Grouping Recommendations AI-generated grouping recommendations help AIOps administrators identify alert correlation opportunities with greater speed and accuracy.

Changed:

Updated supported LLM versions to GPT-4o Mini and Claude Haiku 4.5, bringing the latest model improvements to the platform.

Fixed:

The Autonomous Operator was not automatically updating the description of HLA alerts. This has been corrected.

Fixed - Streamline ACLs

• Fixed ability to add to an exploration a list type data visualization.
• Fixed the ability to edit a data visualization in Visualization Designer.

Updated app/plugin dependencies.

Updated app/plugin dependencies.

• New
  • Upload files to Build Agent to provide context: images (PNG, JPEG, GIF, WEBP), documents (PDF, DOC, DOCX, XLS, XLSX), and text/code files (TS, JS, PY, JSON, MD, HTML, YAML, and more)
  • View update sets created by Build Agent from within the chat panel. Each checkpoint includes a button that opens the relevant update set in a new tab.
  • Connect Build Agent to external MCP servers in ServiceNow Studio. Previously, MCP server connectivity was only available in the ServiceNow IDE.
  • Validate user interface output during app creation with the UI validation tool in Build Agent, now available in ServiceNow Studio.
  • Use the Search retrieval tool to enable agents to fetch and present relevant information from configured data sources in response to user queries. Agents can surface knowledge articles, catalog items, and other indexed content directly within the agentic workflow, reducing the need for users to navigate to separate search interfaces.
• Changed: New MCP integrations in ServiceNow Studio: Added support for many new connectors including Atlassian Rovo, DocuSign, Figma, GitHub, Linear, Miro, Prisma Postgres, and Zoom (Chat, Docs, Revenue Accelerator, Whiteboard)

1. AI Agents and skills

2. Business Continuity management features for impact analysis, business continuity plans, crisis management, and crisis map.

1. AI Agents and skills

2. Business Continuity management features for impact analysis, business continuity plans, and crisis management.

Restrict context based on requestor's data

• Support execution of Agent Chat, Sidebar Chat and Call Summarization to execute through Mosaic
• Restrict Agent Chat Summarization Card only to the agent invoking it
• Provide support to add default tables in Sidebar Chat Summarization configuration
• Uptake Cobalt Raven ACLs Brazil Directive on the table sys_generative_ai_predictions_log

New:

• Team Member Role for SPM Project Workspace - Enables team members to access CWM and manage project tasks. Users can perform CRUD operations on project tasks, demand, and idea tables, and have calendar access. Role assignments and permissions are managed to enhance collaboration within the workspace.
• AI-powered import of Tasks and Stories into CWM Boards from spreadsheets, documents, and images using Now Assist.
• Inline comments in Docs to collaborate on specific text in a document through threaded discussions. Mention colleagues with @ to notify them directly, and receive email notifications when a comment is added, a reply is posted, or you are @-mentioned in a Doc.

New:

AI Framework that autonomously discovers schema, samples, and aggregates live ServiceNow tables, and presents a sourced answer without modifying any data.

Enhancements and defects fixed:

1. AI modified field indicator. AI Sparkle on Fields: AI-generated field indicators are now displayed consistently across both the platform UI and Workspace.
2. Complaint Case Intake Agent
   • Added the hyperlink to the complaint case number.
   • Prompt enhancements on location and products
   • Related Parties for Consumers.
3. Complaint Case Research Agent
   • Deleted default tasks after the creation of agent tasks.
   • Prompt enhancement on multi messages into one.
   • Search profile separation

• Changed
  • Identify missing clauses in contract revisions with improved accuracy.
  • Support for upgraded versions of third party LLMs has been provided.
• Fixed
  • The activity stream now correctly displays the document name when an AI review is initiated, replacing the previous 'null' placeholder.
  • Email notifications to the contract administrator are now reliably sent when metadata extraction is completed.

Performance improvement.

New: Conversation Insights now supports customers with data sovereignty requirements. The application can be re-enabled for customers using local SKUs, allowing deployment with local conversation insights services.

• Changed: None
• Fixed: The assignment of the sn_aci.insights_read_write role to the conv-insights-svc-account user has been corrected to ensure proper handling of inheritance flag information.
• Removed: None

• New
  • General access for Premium Chat on Portals, Employee Slate, MS Teams, and Mobile
  • Voice Input for Premium Chat
  • Voice Assistant support on the 3rd party mobile app.
  • New metrics supported for the conversation auto evaluation feature
• Changed
  • Now Assist Panel Platform (default) assistant would be turned on by default for new customers
  • UX improvement on the Assistant edit flow
  • Inactive assistants can be tested in the Testing panel
  • Assets can be assigned to inactive assistants.
  • Setup topics without Gen AI skill will have visible and discoverable false, toggle will be disabled. For Setup topics that have Gen AI skills, admin can still toggle these flags.
• Fixed
  • Number of Promoted Assets is no longer restricted from the Assistant edit flow
  • Accessibility color contrast

Core Business Suite (CBS) brings together case and knowledge management applications for essential business functions including HR, Source-to-Pay, Finance, Workplace Services, Legal, and Health & Safety. It enables organizations to standardize, manage, and streamline critical employee and business workflows on a single, unified platform.

• Employee Slate is now the default experience - Core Business Suite now uses Employee Slate as the requester experience out-of -the-box, giving employees a single, consistent place to get help.
• CBS Analytics Dashboard - A new dashboard gives a consolidated, cross-business-unit view of service performance, including open and unassigned cases, mean time to resolve, and self-service resolution.
• Core Business Suite Advanced is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Advanced applications.

Core Business Suite Advanced for Human Resources is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Advanced HR applications.

Core Business Suite Advanced for Legal is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Advanced Legal applications.

Core Business Suite Advanced for Source to Pay is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Advanced S2P applications.

Updated to support the latest versions of the dependent apps.

Shipped out-of-the-box assignment rules for the Core Business Suite for Finance. The goal is to give finance admins immediate visibility into existing rule configurations without manual setup, letting them review and understand what rules are active right out of the box.

Shipped out-of-the-box assignment rules for the Core Business Suite for Health and Safety. The goal is to give Health and Safety admins immediate visibility into existing rule configurations without manual setup, so they can review and understand which rules are active by default.

Core Business Suite for Human Resources delivers end_to_end Human Resources case and knowledge management, combined with other enterprise workflows.

Shipped out-of-the-box assignment rules for the Core Business Suite for Legal. The goal is to give legal admins immediate visibility into existing rule configurations without manual setup, letting them review and understand what rules are active by default.

The Core Business Suite For Source To Pay contains all Source To Pay-related code within this applications repository. This application will not be visible to customers on the ServiceNow Store and is intended solely to organize legal-related code within CBS.

Shipped out-of-the-box (OOB) assignment rules for the Workplace Service Delivery administrators. This change surfaces the OOB rules in the admin interface, giving teams an immediate and accurate picture of the baseline configuration.

• Employee Slate is now the default experience - Core Business Suite now uses Employee Slate as the requester experience out-of -the-box, giving employees a single, consistent place to get help. Install and configure it from the CBS admin setup, with support for both Moveworks and Now Assist.
• Assignment rules for all business units - Route cases to the right team automatically with configurable assignment rules across HR, Legal, Finance, Source-to-Pay, Workplace Services, and Health & Safety. Out-of-the-box rules are included for each business unit and can be created or edited from the CBS console.
• CBS Analytics Dashboard - A new dashboard gives a consolidated, cross-business-unit view of service performance, including open and unassigned cases, mean time to resolve, and self-service resolution.

Shipped out-of-the-box assignment rules for the Core Business Suite for Finance. The goal is to give finance admins immediate visibility into existing rule configurations without manual setup, so they can review and understand which rules are active by default.

Shipped out-of-the-box assignment rules for the Core Business Suite for Health and Safety. The goal is to give Health and Safety admins immediate visibility into existing rule configurations without manual setup, so they can review and understand which rules are active by default.

Shipped out-of-the-box assignment rules for the Core Business Suite Foundation for Human Resources. The goal is to give HR admins immediate visibility into existing rule configurations without manual setup, so they can review and understand which rules are active by default.

Shipped out-of-the-box assignment rules for the Core Business Suite for Legal. The goal is to give legal admins immediate visibility into existing rule configurations without manual setup, letting them review and understand what rules are active by default.

Shipped out-of-the-box assignment rules for the Core Business Suite Foundation for Source to Pay. The goal is to give S2P admins immediate visibility into existing rule configurations without manual setup, so they can review and understand which rules are active by default.

Shipped out-of-the-box assignment rules for Core Business Suite for Workplace Services. The goal is to give Workplace Services admins immediate visibility into existing rule configurations without manual setup, letting them review and understand what rules are active right out of the box.

• Employee Slate is now the default experience - Core Business Suite now uses Employee Slate as the requester experience out-of -the-box, giving employees a single, consistent place to get help.
• CBS Analytics Dashboard - A new dashboard gives a consolidated, cross-business-unit view of service performance, including open and unassigned cases, mean time to resolve, and self-service resolution.
• Core Business Suite Prime is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Prime applications.

Core Business Suite Prime for Human Resources is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Prime HR applications.

Core Business Suite Prime for Legal is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Prime Legal applications.

Core Business Suite Prime for Source to Pay is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Prime S2P applications.

Core Business Suite Prime for Workplace Services is now available for direct installation from the Product Hub. This enables one-click installation of all CBS Prime of Workplace Services.

Changed: Maintenance release.

Changed:

• Added voice-driven case status retrieval and updates, enabling customers to check open case statuses and submit updates through guided voice interactions across Genesys, NICE, Five9, 3CLogic, and Amazon Connect, reducing live agent dependency.
• Enhanced Provide Customer 360 Insights with Enterprise Graph and AI Agent deep research for richer, more contextual query results.

• Agent renamed to "Vision AI Agent"
• Requests now routed via the MultiModal Service (MMS) to enable more consistent and scalable multimodal processing (up to 500MB video files)
• Improved response quality for Video Workflow use cases
• Supported video formats: MP4, MOV, WEBM, AVI, MKV, and WMV

Fixed:

Updated the version for ServiceNow Docs Connector to fix dependency failures

This release provides the following enhancements and improvements.

• Enhanced accessibility support Performance improvements Support for L10n Enhanced telemetry support for tracking user behavior Enhanced the home page layout with auto-adjust flexibility based on the widgets count Enhanced Admin configuration to support Tasks and requests (fka Inbox), To-dos task type configurations
• Enhanced AI Preferences to support Smart Prioritisation and AI-Generated Summarisation
• Enhanced the option to load AI summary automatically or on demand on AI summary details page

[Changed]

This application replaces the Now Assist for IRM application to access the AI-powered features.

[Changed]

This application replaces the Now Assist for IRM application to access the AI-powered features.

New

Use the Enterprise Architecture query agent to ask natural language questions about your enterprise architecture portfolio. The agent supports multi-condition queries, quarterly trend comparisons, and impact analysis for scenarios such as application decommissioning, and suggests related questions after answering.

• Generate a business process map (BPM) diagram in the Enterprise Modeling & Visualization by uploading an image of an existing process diagram.
• Use the Enterprise Architecture query agent to ask natural language questions about your enterprise architecture portfolio. The agent supports multi-condition queries, quarterly trend comparisons, and impact analysis for scenarios such as application decommissioning, and suggests related questions after answering.

• New: Improvements to the AI Agent to support NextWave compatible experience.
• Fixed
  • Agent prompt quality improvements.
  • Hallucination issues to improve AI Agent performance.
• Removed: Implicit dependency on the Webcrawler external content connector. The Webcrawler connector now needs to be installed explicitly as a soft dependency.

• AI-powered mobile form auto-fill for field technicians.: Field technicians can use mobile camera to capture images and auto-populate form fields via AI image recognition, supporting smart assessments, input forms, and scripted questionnaires.
• Consistent AI visual indicators across applications.: AI features now display unified gradient styles and subtle animations to indicate activity in Workspace, UI16, and mobile experiences, meeting accessibility standards.
• Parts Debrief automation in NAP and NAVA.: Agents can automate parts debrief workflows, handling empty work notes, invalid part usage, quantity normalization, and error scenarios.
• Expanded AI summarization and knowledge workflows.: Users can trigger work order/sidebar chat summarization and knowledge article generation across platforms with role-based workflows and answer validation.
• Create work order agent with enhanced persona coverage.: System supports creating work orders from text/images in NAP/NAVA with persona validation, blocking invalid content.
• FSM AI Mandate features available on mobile and workspace.: FSM AI capabilities (IT automation, NAVA integration, smart assessments) are now mobile/workspace accessible with role-based access.
• Telemetry for AI feature usage.: Admins can track AI lens feature usage for FSM use cases.
• Skill refinement for FSM digital agents.: FSM agents enhanced for improved automation and workflow handling.
• Changed
  • Broadened test and validation coverage for FSM AI use cases.: Testing expanded across mobile/agent/NextWave experiences to ensure certification and release standards.
  • Enhanced automation for AI SKU and agent studio test cases.: Remaining automation scenarios completed with monitoring on latest NAP version.
  • Improved parts debrief workflow handling.: Better management of edge cases like unrecognized assets and invalid task states.
  • Refined questionnaire answer validation in summarization workflows.: Validation improved across NAP/NAVA/UI16/Workspace for task summarization.
  • Performance testing for Q1 deliveries.: Testing conducted to ensure system stability and responsiveness.

New:

• Added Insurance CSR support AI agent for Interaction page
• Added Insurance CSR customer insights AI agent for Customer 360 Page
• Added Insurance Data Lookup tool
• Added Conversation Transcript Reader tool

• Summarize flow execution details when flow reporting is off.
• Identify potential causes of flow errors.
• Resolve flow errors with suggested fixes.

Minor defect fixes

Minor defect fixes

• AI-powered mobile form auto-fill for field technicians.: Field technicians can use mobile camera to capture images and auto-populate form fields via AI image recognition, supporting smart assessments, input forms, and scripted questionnaires.
• Consistent AI visual indicators across applications.: AI features now display unified gradient styles and subtle animations to indicate activity in Workspace, UI16, and mobile experiences, meeting accessibility standards.
• Parts Debrief automation in NAP and NAVA.: Agents can automate parts debrief workflows, handling empty work notes, invalid part usage, quantity normalization, and error scenarios.
• Expanded AI summarization and knowledge workflows.: Users can trigger work order/sidebar chat summarization and knowledge article generation across platforms with role-based workflows and answer validation.
• Create work order agent with enhanced persona coverage.: System supports creating work orders from text/images in NAP/NAVA with persona validation, blocking invalid content.
• FSM AI Mandate features available on mobile and workspace.: FSM AI capabilities (IT automation, NAVA integration, smart assessments) are now mobile/workspace accessible with role-based access.
• Telemetry for AI feature usage.: Admins can track AI lens feature usage for FSM use cases.
• Skill refinement for FSM digital agents.: FSM agents enhanced for improved automation and workflow handling.
• Changed
  • Broadened test and validation coverage for FSM AI use cases.: Testing expanded across mobile/agent/NextWave experiences to certify features.
  • Enhanced automation for AI SKU and agent studio test cases.: Remaining automation scenarios completed with monitoring on latest NAP version.
  • Improved parts debrief workflow handling.: Better management of edge cases like unrecognized assets and invalid task states.
  • Refined questionnaire answer validation in summarization workflows.: Validation improved across NAP/NAVA/UI16/Workspace for task summarization.
  • Performance testing for Q1 deliveries.: Testing conducted to ensure system stability and responsiveness.

• New
  • AI-powered mobile form auto-fill for field technicians.: Field technicians can use mobile camera to capture images and auto-populate form fields via AI image recognition, supporting smart assessments, input forms, and scripted questionnaires.
  • Consistent AI visual indicators across applications.: AI features now display unified gradient styles and subtle animations to indicate activity in Workspace, UI16, and mobile experiences, meeting accessibility standards.
  • Parts Debrief automation in NAP and NAVA.: Agents can automate parts debrief workflows, handling empty work notes, invalid part usage, quantity normalization, and error scenarios.
  • Expanded AI summarization and knowledge workflows.: Users can trigger work order/sidebar chat summarization and knowledge article generation across platforms with role-based workflows and answer validation.
  • Create work order agent with enhanced persona coverage.: System supports creating work orders from text/images in NAP/NAVA with persona validation, blocking invalid content (links, code, special characters).
  • FSM AI Mandate features available on mobile and workspace.: FSM AI capabilities (IT automation, NAVA integration, smart assessments) are now mobile/workspace accessible with role-based access.
  • Telemetry for AI feature usage.: Admins/analysts can track AI lens feature usage for FSM use cases.
  • Skill refinement for FSM digital agents.: FSM agents enhanced for improved automation and workflow handling.
• Changed
  • Broadened test and validation coverage for FSM AI use cases.: Testing expanded across mobile/agent/NextWave experiences to certify that features meet release standards.
  • Enhanced automation for AI SKU and agent studio test cases.: Remaining automation scenarios completed with monitoring on latest NAP version.
  • Improved parts debrief workflow handling.: Better management of edge cases like unrecognized assets and invalid task states.
  • Refined questionnaire answer validation in summarization workflows.: Validation improved across NAP/NAVA/UI16/Workspace for task summarization.
  • Performance testing for Q1 deliveries.: Testing conducted to ensure system stability and responsiveness.

Fixed : None

Removed : None

Initial release

New: Guardian prompt injection detection is being enabled as always-on and in blocking mode by default for Skills. Added granular admin controls to enable Guardian's prompt injection detection independently for skills, agents, or both

Changed: Increased Gen AI execution log (sys_generative_ai_log) retention to six months.

This release includes improvements to GRC Common Generative AI in issue resolution, issue action planning, agent skill configuration, subscription management, and translations.

New Fetch-Issue-Details Tool: A new tool enables the issue resolution AI agent to fetch issue details by issue number or identifier, complementing the existing tool and enhancing response accuracy. No upgrade action is required, but customers with customizations should review the new mappings.

Refined Action Plan and Remediation Agents: Improvements to the issue action plan and remediation tasks AI agents and their supporting tools ensure more reliable and consistent generation of action plans and remediation suggestions across multiple AI model providers, with better usage tracking. Customers with custom post-processors should re-validate changes.

Updated Issue Summarization Skill: The Now Assist issue summarization skill includes new configuration records offering descriptions and guides for configuration and prompt customization, enhancing the administrator and end-user experience without requiring upgrade actions.

Generative-AI Asset Subscriptions and Translations: Eight new asset subscriptions have been added for tracking and governance of GRC AI capabilities, and translations across supported languages have been refreshed to improve the experience for non-English users. Customers should verify subscription states post-upgrade.

This release includes the following changes:

• Added default deny ACLs on Control Objective skills, the Risk Identification Agent, and the Risk Assessment Summarization skill.
• Excluded audit entry control objectives from skills.
• Enabled audit entry support for related records.
• Excluded audit entry risks from the Risk Identification Agent.
• Enabled third-party model support for the Risk Identification Agent and the Risk Assessment Summarization skill.

New:

Backend support for HR Knowledge Gaps. This capability will be taken up by the Knowledge Center team in a future release. No customer-facing updates in this release.

Changed:

Nothing changed in this release.

Fixed:

Various minor defects fixed.

Removed:

Nothing removed in this release.

In this version, all generated summaries and user-facing content now display in the user's preferred language.

Resolved defects affecting content and agent tool functionality

New: Added new skills for Interview health tracker.

Fixed: AI Agent records view was affected due to an AI Native ACL issue

Changed

App and plugin dependencies have been updated.

Changed

App and plugin dependencies have been updated.

Changed

App and plugin dependencies have been updated.

New: Update to include AI native and risk calculator

New: Update to include AI native and risk calculator

New: Update to include AI native and risk calculator

New:

Updates to support AI Agent Advisor.

Changed:

Nothing changed in this release.

Fixed:

Nothing fixed in this release.

Removed:

Nothing removed in this release.

The IRM Advanced plugin has updated its dependencies by replacing the Now Assist IRM plugin with the IRM Professional plugin, while maintaining dependencies on AI Platform Advanced and IRM Foundation. This change means that installing the Advanced tier now includes all Professional tier features, but generative AI capabilities are no longer included transitively and remain accessible through the Foundation dependency. Customers upgrading to this release will have the IRM Professional plugin added automatically and should verify the status of the IRM Generative AI plugin, deactivating it if preferred.

The IRM Foundation plugin has been updated to include a new dependency on the IRM Generative AI plugin, in addition to the existing dependencies on AI Platform Foundation and IRM Standard. This change ensures that installing or upgrading IRM Foundation automatically includes generative AI capabilities for all IRM tiers built on Foundation, including higher tiers such as Standard, Professional, Advanced, and Enterprise. Upon upgrading, the IRM Generative AI plugin is installed automatically, but customers may deactivate it if they do not wish to use its capabilities. Additionally, customers should verify that AI-related entitlements, such as generative AI subscriptions, are properly assigned to users who will use these features.

The IRM Prime plugin now depends on IRM Enterprise, in addition to AI Platform Foundation and IRM Advanced, for installation or activation. This change ensures that AI features in Prime rely on data managed by IRM Enterprise entitlements, improving functionality by automatically resolving and including IRM Enterprise if not already installed. Customers upgrading to version 22.3.0 must have IRM Enterprise licensed and available to complete the upgrade; no data migrations or configuration changes are needed, and existing integrations remain unaffected.

New

Implemented meta data changes to enable AI native support for all skills and agents in the Regulatory Change Management application.

Changed

Improved agent instructions and supporting tools for the Regulatory Alert Analysis and Recommendations agent to ensure more reliable and consistent responses.

Fixed: Risk Event Summary now generates correctly when a task is associated with a Risk Event.

New:

Employee experience

Front door employee experience (Employee Slate): a conversation-first portal with a unified Activity Hub displaying open incidents and requested items, fulfillment progress, a Needs Attention area, and a simplified Request experience.

Fulfiller experience

Catalog tasks on the landing page: IT agents can track catalog tasks alongside incidents and requests on the fulfiller homepage.

Admin experience

• Expand to Service Operations: a guided upgrade path adds Major Incident, Problem, On call, and Walk up Management, with one-click install and direct navigation to admin setup.
• Employee Slate IT Service Management onboarding: the configuration console includes the Employee Slate setup task as part of IT Service Management onboarding.
• CMDB onboarding: the configuration console includes configuration item setup task as part of IT Service Management onboarding.

Changed:

Fulfiller experience

• Embedded Now Assist enhancements: the incident record card now surfaces contextual actions at resolution: draft or attach a knowledge article, propose a major incident, create change or problem records, and share resolution steps with the requester. Actions appear only when relevant to the incident context.
• Incident form layout: updates to the Incident details layout, including the position of Attachments and Configuration Items on the form.

Admin experience

• AI agent improvements: improvements to the AI agents that help configure routing rules, categories, and subcategories.
• Configuration console setup tasks: updated header and footer for a consistent experience and clearer understanding of each task.

New:

Fulfiller actions from an incident: Fulfillers can create a change request, propose a major incident, and create a problem record for an incident.

Changed:

• Enhanced embedded Now Assist chat module card with contextual resolution actions: draft or attach a knowledge article, propose a major incident, create a change or problem, and send next steps to the requester.
• AI-assisted incident triage, resolution guidance, and summaries surfaced natively on the record.

1. The L1 Service Desk AI Specialist now detects knowledge gaps and automatically creates a Knowledge Feedback Task when KB search returns no results above the configured relevance threshold, enabling managers to create missing articles.
2. Knowledge feedback tasks are now deduplicated the system checks for existing open tasks with matching category and subcategory within a lookback window before creating a new one, preventing task noise for gaps like password resets.
3. The Change Quality Agent now persists quality scores per change record including timestamps and contributing factors available for dashboards, audits, and risk-detection workflows.
4. A new KB article usage section on the AI Agent Studio Performance dashboard shows which articles the AI Specialist used during autonomous resolution.
5. The AI Specialist now persists the Skill applied per incident evaluation onto the record. A new Skill filter on the AIS Performance Dashboard lets managers segment performance by capability area.
6. The Insights and Opportunities for Incident dashboard in Service Operations Workspace aggregates incidents into trend categories with AI-generated summaries, SLA performance, sentiment, channel adoption, and geographic distribution, giving incident managers a consolidated view of incident patterns with drill-down views into individual records.
7. IT administrators can now configure Change Management through an AI-native conversational agent in the Product Console, guided through approvals, risk scoring, and workflows via natural language.
8. The ZTS AI L1 Service Desk Specialist now supports real-time conversational engagement via Microsoft Teams or Slack, replacing the asynchronous activity stream with a consent-driven experience.
9. A new agentic solution for SharePoint access issues autonomously handles L1 scenarios including access requests, file failures, and permission mismatches, reducing resolution time from hours to minutes.

1. The AI Specialist Performance Dashboard now includes revised productivity metrics measuring incidents attempted against total count for a more accurate resolution baseline.
2. Targeted enhancements to the L1 Service Desk AI Specialist improve reliability and consistency based on quality feedback from recent releases.
3. The Create Incident AI Agent has been migrated to a Hierarchical Agent model, resolving hallucination and rendering issues on NextWave and removing the VA dependency.
4. The Create Incident AI Agent has been updated for full NextWave off-glide orchestrator compatibility, resolving functional deviation issues from GA readiness testing.

• Fixed- None
• Removed - None

• New: n/a
• Changed: Security related enhancements on existing AI Voice agents
• Fixed: n/a
• Removed: n/a

New in ITOM - Advanced

• Added AI agents to automate monitoring and remediation for Discovery, MID Servers, Certificate Management, Service Mapping, and LEAP.
• Added AI Agent Topology Mapping to provide governance and visibility to your AI estate.
• Added ITOM URL Discovery to surface shadow SaaS and unmanaged applications directly in the CMDB.
• Added ITOM Configuration Console to simplify onboarding for Discovery and Event Management with guided, best-practice setup.
• Added AIOps Learning Enhanced Automation Playbooks to turn historical incidents into dynamic, reusable resolution intelligence.

See the Now Assist for Platform and AIOps Experience release notes.

• Service Mapping supports the Model Context Protocol (MCP), providing a conversational interface for live application service data. Service Mapping admins can connect Claude Desktop to a ServiceNow instance and query service topology, server-to-service relationships, related CIs, and unmapped configuration items.
• Service Mapping Workspace widgets1. Business app linked to App services - Indicating the count of business applications that were linked to application services by the Business App Mapping AI Agent.2. Maps created by Now Assist - Indicating the count of the service maps created by Service Mapping AI Agent.
• Changed
  • The AI specialists for Service Mapping are renamed as AI Agents for Service Mapping.
  • View the service map created by Service Mapping AI Agent through the AI Activity list.

1. The Change Quality Agent now persists quality scores in a dedicated table linked to each change record, enabling persistent reporting and downstream use. Scores include metadata such as timestamps, scoring version, and contributing factors, and are available for dashboards, audits, and risk-detection workflows across the full change lifecycle.
2. The Insights and Opportunities for Incident dashboard in Service Operations Workspace aggregates incidents into trend categories with AI-generated summaries, SLA performance, sentiment, channel adoption, and geographic distribution, giving incident managers a consolidated view of incident patterns with drill-down views into individual records.
3. IT administrators can now configure Change Management through an AI-native conversational agent in the Product Console, replacing complex admin navigation. The guided agent walks administrators through key settings including approvals, risk scoring, and workflows via natural language, making Change Management configuration accessible to customers with limited ServiceNow expertise.

1. The Create Incident AI Agent has been migrated from VA topic-based orchestration to a Hierarchical Agent model, resolving hallucination, functional deviation, and rendering issues on the NextWave orchestrator. The agent now runs natively on NextWave without VA dependency, delivering a consistent and reliable incident creation experience for end users.
2. The Create Incident AI Agent has been evaluated and updated for compatibility with the NextWave off-glide orchestrator. All hallucination and functional deviation issues identified during GA readiness testing have been resolved, ensuring consistent and accurate incident creation behavior across both VA-based and NextWave runtime environments.

1. An issue has been resolved where Platform skills were not appearing in the Now Assist Skills admin panel even when the ignoreFulfillerSubscriptionCheck system property was enabled. Administrators can now successfully view and activate Platform skills from the Now Assist admin interface.
2. Several display and interaction issues in the Recommended Actions panel have been resolved, including a misaligned loading indicator on the search box, a non-functional full-view search icon, filter dropdowns obscured by the footer, and an incorrect hand cursor appearing on non-clickable KB article and AI action content.

1. The Create Incident AI Agent has been migrated to a Hierarchical Agent model, resolving hallucination and rendering issues on NextWave and removing the VA dependency.
2. The Create Incident AI Agent has been updated for full NextWave off-glide orchestrator compatibility, resolving functional deviation issues from GA readiness testing.

1. Platform skills not appearing in the Now Assist Skills admin panel (even with ignoreFulfillerSubscriptionCheck enabled) have been resolved.
2. Display and interaction issues in the Recommended Actions panel have been resolved: misaligned loading indicator, non-functional search icon, footer-obscured filters, and incorrect cursor on non-clickable content.

1. The L1 Service Desk AI Specialist now detects knowledge gaps and automatically creates a Knowledge Feedback Task when KB search returns no results above the configured relevance threshold, enabling managers to create missing articles.
2. Knowledge feedback tasks are now deduplicated the system checks for existing open tasks with matching category and subcategory within a lookback window before creating a new one, preventing task noise for gaps like password resets.
3. The Change Quality Agent now persists quality scores per change record including timestamps and contributing factors available for dashboards, audits, and risk-detection workflows.
4. A new KB article usage section on the AI Agent Studio Performance dashboard shows which articles the AI Specialist used during autonomous resolution.
5. The AI Specialist now persists the Skill applied per incident evaluation onto the record. A new Skill filter on the AIS Performance Dashboard lets managers segment performance by capability area.
6. The Insights and Opportunities for Incident dashboard in Service Operations Workspace aggregates incidents into trend categories with AI-generated summaries, SLA performance, sentiment, channel adoption, and geographic distribution, giving incident managers a consolidated view of incident patterns with drill-down views into individual records.
7. IT administrators can now configure Change Management through an AI-native conversational agent in the Product Console, guided through approvals, risk scoring, and workflows via natural language.
8. The ZTS AI L1 Service Desk Specialist now supports real-time conversational engagement via Microsoft Teams or Slack, replacing the asynchronous activity stream with a consent-driven experience.
9. A new agentic solution for SharePoint access issues autonomously handles L1 scenarios including access requests, file failures, and permission mismatches, reducing resolution time from hours to minutes.

1. The AI Specialist Performance Dashboard now includes revised productivity metrics measuring incidents attempted against total count for a more accurate resolution baseline.
2. Targeted enhancements to the L1 Service Desk AI Specialist improve reliability and consistency based on quality feedback from recent releases.
3. The Create Incident AI Agent has been migrated to a Hierarchical Agent model, resolving hallucination and rendering issues on NextWave and removing the VA dependency.
4. The Create Incident AI Agent has been updated for full NextWave off-glide orchestrator compatibility, resolving functional deviation issues from GA readiness testing.

1. Platform skills not appearing in the Now Assist Skills admin panel (even with ignoreFulfillerSubscriptionCheck enabled) have been resolved.
2. Display and interaction issues in the Recommended Actions panel have been resolved: misaligned loading indicator, non-functional search icon, footer-obscured filters, and incorrect cursor on non-clickable content.

The Knowledge Center centralized hub for Knowledge Admin and Manager:

Changes:

Added changes for better Typo handling for Table Labels.

Added internal usage tracking enhancements.

• Discover and execute Ansible playbooks from LEAP automation opportunities to promote faster incident resolution
• Track missed automation opportunities in LEAP to identify gaps in automation coverage
• Auto-generate resolution steps for all critical and high severity automation opportunities post initial run
• Improved resolution step relevance as irrelevant steps filtered out by group problem description
• LEAP available with Now Assist for ITSM

Changed

• Product renamed from AIOps LEAP to LEAP (Learning-Enhanced Automation Platform)
• AI-enhanced and structured knowledge base articles published to the Knowledge Center
• Fixed
  • Automation Opportunities page load time reduced from 3+ seconds
  • Value Dashboard page load time reduced from 4+ seconds
  • Resolved stale resolution steps reappearing in record metric table after SOW Flow updates
  • Fixed sharp-edged AI Gradient button for Regenerate in AO Details to match default theme
  • Fixed UI issues on resolution steps filter in LEAP Settings page
  • Fixed missing help content on LEAP Installer Skill Activation page
  • Reverted unintended default UI filter that auto-generated resolution steps
  • Fixed Maintenance ACL Roles being overridable by Admins in LEAP Automation Playbooks

• Changed: Support for upgraded versions of third party LLMs has been provided.
• Fixed: Legal request and matter summarization now delivers improved accuracy across all supported models

AI Filter Assist allows users to define conditions using natural language in an intuitive, dialog-driven interface. This feature helps reduce the cognitive load of building conditions using the traditional condition builder by allowing users to type or speak a prompt that will then create the query needed to filter, sort, or group data.

This release we fixed two defects in AI Filter Assister where an ACL had a typo in it's script and "Add to existing filters" mode was selectable even though it should be disabled.

• New: Delivery note upload for quantity disputes: Customers can upload delivery notes during case creation, enabling the invoice dispute intake assistant AI agent to validate and resolve quantity disputes without human intervention.
• Changed: Order line quantity validation for disputes: The invoice dispute support assistant AI agent uses order line data when sold product records are unavailable such as new orders, non-serialized products, or services. It auto-approves valid disputes and escalates only when data is unclear.

• New: Automated email notifications for order cases: Send emails when the AI agent opens an order case from a Business Portal chat and when the case is resolved. Closure emails include resolution and quote details, if generated. Voice-created cases dont trigger emails.Scripted extension point for quote thresholds: A scripted extension point to integrate the manage order operations AI agent with your inventory, ERP, and quote systems for quote threshold evaluation. The AI agent generates a quote when a quantity-change request exceeds the configured price threshold.
• Changed
• Order exception support for quantity and shipping location: Customers can request quantity and shipping location updates, along with expedite requests via chat and voice channels.
• Voice assistant intake for order exceptions: Customers can submit expedite, quantity, and shipping location requests via voice. Requests are captured and converted into order cases for resolution in CSM/FSM Configurable Workspace.
• Unified extension point for feasibility checks: A single scripted extension point (sn_ord_ops_aias.orderExceptionCheckEP) validates delivery, quantity, and shipping feasibility, replacing the earlier expedite-only ATP check.
• Auto-populated account and contact on interactions: Account and Contact fields are populated during chat handoff, giving agents immediate customer context and improving summarization accuracy.
• AI-generated summaries for agent handoff: Agents receive a concise AI-generated summary instead of full chat history, improving context and productivity.

New

Detect claim anomaly - two new capabilities

Auto-categorize Repair Claims capability

GenAI summary on repair claims

New:

Access Strategic Portfolio Management data and Now Assist AI skills as MCP tools, enabling LLM agents to query and reason about goals, portfolio plans, and projects. The following tools are available in this release: Tool Description get_goals Retrieves goals and objectives. generate_goal_insights Generates AI-powered insights for goals and targets. get_portfolio_plans Retrieves portfolio plans. generate_portfolio_insights Generates portfolio insights, including at-risk projects, delayed starts and ends, and dependencies. get_projects Retrieves projects. generate_project_insights Detects project risks, analyzes status trajectory, and provides recommendations. identify_project_risks Detects AI-identified RIDAC risks and saves them to the risk table as AI drafts. get_ai_status_report Generates a RAG (Red, Amber, Green) status report across resources, cost, schedule, and scope.

Removed : api-version query parameter from Create Response OEM action

New

Enabled the mobile card generation capability to be executed Off-Glide to improve scalability and performance.

Changed

Minor API changes for another app support internally

Performance improvements implemented, and third-party IDPs now allowed to be used.

New:

• Multi-intent identification support
• Missing inputs extraction
• email branding/template support for email generation

Instance and system administrators can now archive custom skills directly from the Now Assist Skills page. Archiving is not supported for ServiceNow out-of-box (OOB) skills. Archived skills can be restored at any time.

View Detail button is now available alongside the Activate button for OOB skills on the Now Assist Skills page. The Skill detail view provides administrators and users with key information about a skill including its benefits, prerequisites, recommended model provider, and related skills enabling informed activation decisions.

Prompt injection protection can now be configured at both the instance level and the product level (CSM, ITSM, and others), giving administrators more granular control over enforcement across workflows.

Administrators can now launch a guided tour walkthrough directly from the " Manage Model Versions" section within Settings section in now assist admin. The guided tour provides a step-by-step introduction to the model version management workflow, helping admins understand version lifecycle, deprecation alerts, and available actions.

Approval checklist generator skill accepts additional context parameter that is used to resolve checklists unknown items.

• New: AI Agent Extensibility: Customers can now add / remove AI agents or tools from OOTB agents while remaining on the upgrade path.
• Changed: Orchestrator defaults to GPT-5.4: The orchestrator now uses GPT-5.4 as its default model when Azure OpenAI is the selected LLM.
• Fixed
  • Agent Execution
    • Published version instructions not being applied to agents at runtime.
    • Execution plan fetch timing out on instances with large numbers of execution plan records.
    • Agentic workflow stuck at orchestrator step intermittently.
    • VA transaction timeout and conversation stuck on "Thinking."
  • Agent-to-Agent (A2A)
    • Async handling of working state from secondary agent causing primary flow issues.
    • Raw rich control output in agent response causing conversation to fault and close abruptly.
    • input-required response sending empty message/parts block.
  • Multilingual
    • Tool names, agent names, and "Processing new information" untranslated in AI steps.
    • Words not translated when user language differs from system language (intermittent).

• Browser tabs that open during goal execution in adaptive desktop actions remain open after the goal completes. Use the keep_tab_open system property to turn this behavior on or off. The property is turned on by default.
• Adaptive desktop actions is enhanced to improve execution efficiency.

• Analytics dashboards pertaining to AI agents, AI assistants, and skills are consolidated inside Now Assist Center > Monitor.
• Executions dashboard in Now Assist Center: Deep dive into individual executions of assistants and agents to improve performance.
• Business value dashboard in Now Assist Center: Analyze the impact of AI assets on time and cost savings and set formula for calculating value per AI asset (skills, AI agents, agentic workflows)

• AI asset inventory. A new Asset Inventory page provides a single, consolidated view of every AI asset deployed across the instance including AI agents, agentic workflows, skills, subflows, actions, data assets, virtual assistants, topics, catalog items, and knowledge graphs. Each asset type surfaces active/inactive status, and the list supports column personalization so administrators can tailor the view to their needs. Assets from AI Agent Studio, Data Kit, and the platform catalog are unified in a single browsable surface.
• AI readiness assessment. Administrators can now validate their instance's readiness for AI adoption directly within Now Assist Center. The new AI Readiness page supports on-demand and scheduled assessment execution across both Now Assist and Agentic AI assessment types. Results are surfaced in a dedicated Assessment Details page with historical tracking, enabling teams to identify configuration gaps, measure readiness over time, and accelerate the path to their first production AI solution.
• AI Agent Advisor enhancements. The AI Agent Advisor experience has been refined based on feedback from the May release. Updates include an improved automation opportunities list page with better filtering and layout, redesigned Agent Advisor cards, a refreshed home page CTA experience that surfaces opportunities at a glance, a new Agent Advisor settings page, and stack metrics page improvements. These changes reduce friction in the discovery-to-deployment journey and improve the quality of agent recommendations surfaced to administrators.
• Expanded OOTB skill configuration. The conversational administration experience now surfaces additional out-of-the-box skill configuration paths, including skill availability configuration and display experience settings extending the reach of guided, in-product setup to more of the Now Assist skill catalog.

Defects fixes and off glide LLM execution migration

Increased sample data input limit for SDG requests Users (Data Kit Admin) can now select up to 50 seed records from a data collection or table when submitting a Synthetic Data Generation request, up from the previous default of 3. The maximum sample size is also configurable by NADK Admins via sys_property.

Plausibility Metrics for Synthetic Data Introduced plausibility scoring as a new data quality insight layer within the Generated Data view (Data Kit > Home > Synthetic Data tab > Open Generated Data > Data Insight > Plausibility). Validates that generated field values are realistic, internally consistent, and free of garbage data.

OOB data sets

New:

• Leveraging Intent to Action Agentic Workflow for emails to be parsed and categorized as the correct intent to create a invoice inquiry case record.
• Enhanced workflow to enable requesters to accept/reject the case resolution.

Changed: AI skill execution is now performed off glide

Changed: Updated versions of dependent apps.

Now Assist for code generation features now inherit skill-level ACLs configured in Now Assist Admin, ensuring restricted features are only available to authorized users.

New: AI-powered import of Tasks and Stories into CWM Boards from spreadsheets, documents, and images using Now Assist.

Changed: No new features, enhancements or fixes. Since there are updates to the AI agents, the version of this application must also be updated.

Fixed: Security fixes

Existing feature: with latest third party model versions

• New
  • Build Agent
    • Upload files to Build Agent to provide context: images (PNG, JPEG, GIF, WEBP), documents (PDF, DOC, DOCX, XLS, XLSX), and text/code files (TS, JS, PY, JSON, MD, HTML, YAML, and more)
    • View update sets created by Build Agent from within the chat panel. Each checkpoint includes a button that opens the relevant update set in a new tab
    • Connect Build Agent_to external MCP servers in_ServiceNow Studio. Previously, MCP server connectivity was only available in the_ServiceNow IDE.
    • Validate user interface output during app creation with the UI validation tool in_Build Agent, now available in_ServiceNow Studio.
    • Use the Search retrieval tool to enable agents to fetch and present relevant information from configured data sources in response to user queries. Agents can surface knowledge articles, catalog items, and other indexed content directly within the agentic workflow, reducing the need for users to navigate to separate search interfaces.
  • Catalog Item Generation: Migrated Now Assist for Catalog Generation skill group (Create Catalog Item and Refine Skill) to Mosaic (off-Glide GenAI Controller) by enabling offglide routing via OneExtend
  • Flow Execution Analysis: Net new app in Now Assist for Creator
  • Playbook Generation: Playbooks now automatically generate a concise, AI-powered summary of a Playbook's purpose, structure, and logic, helping new process owners quickly understand existing workflows without manual documentation.
  • Widget Generation and Widget Updation: Australia support
• Changed: Build Agent <ul><li>New MCP integrations in ServiceNow Studio: Added support for 12 connectors including Atlassian Rovo, DocuSign, Figma, GitHub, Linear, Miro, Prisma Postgres, and Zoom (Chat, Docs, Revenue Accelerator, Whiteboard)</li></ul>

New

Dependency on store app Now Assist for CSM Major Issue Management

Changed

Engagement tool for DEX agents to obtain employee approval before auto-triggering device actions. This change integrates with ZTS so will not be published until ZTS is GA.

Updated to support the latest version of the dependent apps.

• Use the Enterprise Architecture query agent to ask natural language questions about your enterprise architecture portfolio. The agent supports multi-condition queries, quarterly trend comparisons, and impact analysis for scenarios such as application decommissioning, and suggests related questions after answering.
• Generate a business process map (BPM) diagram in the Enterprise Modeling & Visualization by uploading an image of an existing process diagram.

Initial Release

• AI-powered mobile form auto-fill for field technicians.: Field technicians can use mobile camera to capture images and auto-populate form fields via AI image recognition, supporting smart assessments, input forms, and scripted questionnaires.
• Consistent AI visual indicators across applications.: AI features now display unified gradient styles and subtle animations to indicate activity in Workspace, UI16, and mobile experiences, meeting accessibility standards.
• Parts Debrief automation in NAP and NAVA.: Agents can automate parts debrief workflows, handling empty work notes, invalid part usage, quantity normalization, and error scenarios.
• Expanded AI summarization and knowledge workflows.: Users can trigger work order/sidebar chat summarization and knowledge article generation across platforms with role-based workflows and answer validation.
• Create work order agent with enhanced persona coverage.: System supports creating work orders from text/images in NAP/NAVA with persona validation, blocking invalid content.
• FSM AI Mandate features available on mobile and workspace.: FSM AI capabilities (IT automation, NAVA integration, smart assessments) are now mobile/workspace accessible with role-based access.
• Telemetry for AI feature usage.: Admins can track AI lens feature usage for FSM use cases.
• Skill refinement for FSM digital agents.: FSM agents enhanced for improved automation and workflow handling.

Changed

• Broadened test and validation coverage for FSM AI use cases.: Testing expanded across mobile/agent/NextWave experiences to certify features.
• Enhanced automation for AI SKU and agent studio test cases.: Remaining automation scenarios completed with monitoring on latest NAP version.
• Improved parts debrief workflow handling.: Better management of edge cases like unrecognized assets and invalid task states.
• Refined questionnaire answer validation in summarization workflows.: Validation improved across NAP/NAVA/UI16/Workspace for task summarization.
• Performance testing for Q1 deliveries.: Testing conducted to ensure system stability and responsiveness.

New:

• Added Insurance Customer Profile Summarization skill for Customer 360
• Added Insurance interaction context summary skill

Minor defect fixes

This release delivers UI modernization and quality improvements. The Now Assist for HAM interface has been upgraded to the latest UI framework, with improved parsing and rendering of Asset Summary responses. Accessibility improvements have also been applied.

Improvements:

1. Currency fields extracted from purchase order attachments now display correctly.
2. A missing field configuration in the purchase order attachment flow has been added.
3. A component reliability issue in the record review workflow has been fixed.

New:: Contextual Action Planner sidebar in the Health and Safety Workspace: Create actions manually or from AI suggestions, and review them in context across incidents, observations, investigations, cases, audits, work permits, risk assessments, meetings, etc.

Consistent model support and on prem support

Security guardrails and Defect fixes related to content and metrics

Gen AI Trace logs improvement

Code Fix Agent: Enhanced Error Logging

When the Code Fix Agent runs into a problem, you will now see a plain-language explanation of what went wrong and what to do next right on the remediation record. Select the links navigate directly to the relevant error record, so there's no need to hunt through system logs.

The Now Assist for IRM application has been reclassified as an installed-as-dependency plugin, making it a platform-managed dependency that is hidden from the main user interface and no longer appears as a standalone application tile in the ServiceNow App Store. Despite this change, AI-driven IRM functionality remains unchanged and continues to operate as before. Users should now access AI features through the IRM Native SKU applications; the AI dependency loads transparently, and no customer action is required to maintain existing functionality.

New:

• AIOps AI Specialist Autonomous Alert Management (Limited Availability) The AIOps AI Specialist is ServiceNow's purpose-built AI agent for IT Operations. It autonomously monitors, triages, and resolves infrastructure and application alerts, executing the full NOC workflow without requiring human intervention on every event. In this release, the AI Specialist can:
  • Automatically pick up alerts based on configurable scope and filter rules
  • Run alert analysis, investigation, and impact assessment autonomously
  • Collaborate with the SRE Agent for deep-level investigation using A2A integration with observability tools and hyperscalers
  • Hand off to a remediation agent to execute subflow-based fixes
  • Close alerts determined to be insignificant noise
  • Be triggered manually by assigning an alert directly to the AI Worker user
• Onboarding Experience for the AIOps AI Specialist A guided setup flow helps NOC team managers configure and activate the AI Specialist quickly and consistently.
• Remediation Agent A new remediation agent works alongside the AIOps AI Specialist to automatically identify and execute the appropriate remediation subflow for an alert.
• Proactive Grouping Recommendations Alert Assist now proactively analyzes incoming alerts and recommends grouping candidates before a NOC operator opens a case. The engine uses intelligent field selection to identify the most relevant alert attributes, supports all LLM types, and handles high-cardinality and verbose alert data gracefully. Recommendations are refreshed every 5 minutes via a scheduled job.
• AI Resolution Attribution When the AIOps AI Specialist closes an alert, the system records it as "Closed by AI," capturing which alerts were fully handled autonomously. This enables NOC teams to measure automation rate and track AI-driven closure trends over time.
• Text to Configuration Agent Available out of the box, the Text to Configuration Agent allows operators to configure AIOps settings using natural language no form-filling required.

Changed:

• The AI Worker alert management rule now supports coexistence with autonomous mode.
• The auto-close path for insignificant noise alerts has been updated for more reliable handling.
• Autonomous Styled Output now generates a remediation anchor to support downstream remediation workflows.
• LLM Model Upgrades Alert Assist skills have been updated to use gpt-4o-mini (previously gpt-small) and Claude Haiku 4.5 (previously claude-small), improving output quality and aligning with current model availability.

Fixed:

• The AI Investigation Report displayed an infinite loading spinner on timeout.
• The Regenerate action in the AI Insights card was non-deterministic. The Express List also showed incorrect insight data when the first analysis run failed.
• The Autonomous Operator was not updating descriptions on individual HLA alerts.
• Alert Assist was logging unnecessary errors when no related changes passed the relevance threshold.
• Alert Assist was logging null pointer exceptions during skill availability checks.

1. The L1 Service Desk AI Specialist now detects knowledge gaps and automatically creates a Knowledge Feedback Task when KB search returns no results above the configured relevance threshold, enabling managers to create missing articles.
2. Knowledge feedback tasks are now deduplicated the system checks for existing open tasks with matching category and subcategory within a lookback window before creating a new one, preventing task noise for gaps like password resets.
3. The Change Quality Agent now persists quality scores per change record including timestamps and contributing factors available for dashboards, audits, and risk-detection workflows.
4. A new KB article usage section on the AI Agent Studio Performance dashboard shows which articles the AI Specialist used during autonomous resolution.
5. The AI Specialist now persists the Skill applied per incident evaluation onto the record. A new Skill filter on the AIS Performance Dashboard lets managers segment performance by capability area.
6. The Insights and Opportunities for Incident dashboard in Service Operations Workspace aggregates incidents into trend categories with AI-generated summaries, SLA performance, sentiment, channel adoption, and geographic distribution, giving incident managers a consolidated view of incident patterns with drill-down views into individual records.
7. IT administrators can now configure Change Management through an AI-native conversational agent in the Product Console, guided through approvals, risk scoring, and workflows via natural language.
8. The ZTS AI L1 Service Desk Specialist now supports real-time conversational engagement via Microsoft Teams or Slack, replacing the asynchronous activity stream with a consent-driven experience.
9. A new agentic solution for SharePoint access issues autonomously handles L1 scenarios including access requests, file failures, and permission mismatches, reducing resolution time from hours to minutes.

1. The AI Specialist Performance Dashboard now includes revised productivity metrics measuring incidents attempted against total count for a more accurate resolution baseline.
2. Targeted enhancements to the L1 Service Desk AI Specialist improve reliability and consistency based on quality feedback from recent releases.
3. The Create Incident AI Agent has been migrated to a Hierarchical Agent model, resolving hallucination and rendering issues on NextWave and removing the VA dependency.
4. The Create Incident AI Agent has been updated for full NextWave off-glide orchestrator compatibility, resolving functional deviation issues from GA readiness testing.

1. Platform skills not appearing in the Now Assist Skills admin panel (even with ignoreFulfillerSubscriptionCheck enabled) have been resolved.
2. Display and interaction issues in the Recommended Actions panel have been resolved: misaligned loading indicator, non-functional search icon, footer-obscured filters, and incorrect cursor on non-clickable content.

• Changed: Support for upgraded versions of third party LLMs has been provided.
• Fixed: Legal request and matter summarization now delivers improved accuracy across all supported models

New

Detect claim anomaly - two new capabilities

Auto-categorize Repair Claims capability

GenAI summary on repair claims

Changed

This release includes security enhancements that strengthen access controls across the application.

Fixed: - ESG business domain scope and knowledge graph tags for disclosure report Corrected the ESG business domain scope and updated knowledge graph tags for the ESG disclosure report to ensure accurate AI-assisted content generation.

Automate the end-to-end creation, validation, and resolution of order exception cases on the Business Portal for Expedite, Additional Quantity, and Ship to Different Location requests.- Capture customer inputs in the Virtual Agent, extract intent and values inline, validate them against inventory, availability, location, and unit-of-measure rules, and confirm with the customer before acting.- Run a pricing threshold check and, when required, generate a quote for customer approval before resolving the case and updating case lines to Resolved.- Hand off seamlessly to a live agent at any failure or rejection point, with full conversation context, case number, and a clickable case link in NAVA preserved.

Updated App/Plugin dependencies

Updated app/plugin dependencies

Updated app/plugin dependencies

Updated app/plugin dependencies

• New
  • Implemented meta data changes to enable AI native support to all the skills and agents on regulatory change management product.
  • All the plugin dependency & role changes to enable AI Native capabilities are implemented as part of this application.
• Changed: The Now Assist issue summarization skill includes new configuration records offering descriptions and guides for configuration and prompt customization, enhancing the administrator and end-user experience without requiring upgrade actions.

New:

• Total session time: Measures the complete end-to-end task duration, including AI processing, tool usage, and user input delays.
• Agent processing time: Captures the time the AI agent spends working on the task, not counting user delays or waiting.
• Customers will be able to view latency numbers in evaluation run reports (Agentic Evaluation). Two latency scores will be available:
  • Agent Instruction Optimization - added support for the latest third-party (3P) models, Hierarchical Agents, and Voice Agents.
  • 'Suggested Fixes' in Agentic Evaluation - added support for Hierarchical Agents.
  • Added Skill Group capability to update LLM providers across all Agentic Judges simultaneously through Now Assist Admin.
• Changed
  • Modified Overall Task Completeness evaluation for improved evaluation accuracy.
  • Agent Instruction Optimization - Enhanced gatekeeping mechanisms for AIO Log Analyzer, Agent Instruction Generator, Tool Description Generator

NOTE: These features and additions will be usable within Now Assist Skill Kit > Agentic Evaluation

Multilingual Enhancements

ICM Case Narrative Ai Agent Support

New:

The automatic conversion of supplier emails into PO exceptions workflow now supports:

• Additional languages, enabling better recognition of non-English supplier communications
• ERP purchase order IDs, allowing precise matching with purchase order data from enterprise resource planning systems

• Fallout creation AI Agent - This will be automatically generate fallouts in the correct category depending on the Notes/Comments on the Task.
• Document to Task plan template AI Agent - This Agent will be used to create a Task plan template using an document ( pdg, png,jpeg) which contains the fulfilment view of the Specification.

New dependant plugin - Opportunity Management AI Features added.

• Submitting feedback on AI-generated recommended actions fails for non-admin users (e.g., analysts).
• Correlation Insights - No output for certain CIs.
• Generate Insights action visible even when Now Assist for Security Incident Response was not installed.
• Changed
  • Summarization skill enhancement to honor chronological order of events and to give more weightage to recent activity.
  • AI Gradient styling for Correlation Insights and Summarization.

Fixed: Adoption of AI Gradient UI for integration builder.

Connections tab in the Service Exchange Center

Create, view, request, and offboard provider and consumer connections from a single location in the Service Exchange Center. Search and filter connections without navigating across multiple screens.

Improved consumer registration and onboarding

Onboard consumers faster with a guided, step-by-step registration experience. Upgraded consumers are automatically redirected to this experience to receive clearer progress indicators during onboarding, and actionable messaging for failure and delay scenarios, minimizing onboarding friction and support dependency.

Improved FDS capabilities

• Improve your connection experience, by syncing Knowledge Base articles between provider and consumer instances.
• Reduce data inconsistencies by maintaining sys IDs for CMDB data and dependent relationships through transform maps.
• Ensure CI functionality is preserved on the destination instance by choosing to automatically create CI dependency relationships when relationship data is received from the source.
• Improved compliance through restricted data sync from non production instances to production instances for CMDB tables.

Journal Field Framework enhancements

• Increase flexibility in journal data synchronization between provider and consumer instances by mapping multiple source fields to a single target journal field.
• Configure journal fields such of type journal_input fields alongside journal type, ensuring all journal entries are preserved during synchronization without requiring custom scripting

Group-based persona assignments for Remote Catalog

Assign Remote Catalog personas to user groups so existing group-based access management practices extend to Remote Catalog, reducing administrative effort by managing access at the group level instead of individual users.

• Tiered SKU Support for AI Native ITSM & CBS: Added support for Advanced and Prime SKUs across AI Native ITSM and CBS, including associated entitlement-driven flows and experiences.
• Product Hub UX Enhancements: Introduced targeted UX improvements to Product Hubs, including updates to the Resources section and guided content refinements based on leadership and customer feedback to create a more intuitive onboarding experience.
• LitJS Component Adoption in NAP:Enabled the ability to import and leverage LitJS components within NAP, helping establish a more scalable and consistent component framework.
• Admin Home Banner Updates: Updated Admin Home banners and messaging to better guide users through setup, actions, and key implementation milestones.
• Zero-Touch Implementation: Expanded Zero-Touch Implementation capabilities to deliver provisioned instances with entitled plugins and apps fully enabled, giving customers a value-ready environment from day one.
• Intelligent Configuration Update Tracking: Built on the April GA release with smarter tracking of configuration updates and changes, helping customers stay informed of new capabilities and recommended actions.
• Targeted UI Improvements: Delivered incremental UI enhancements across the experience, informed directly by customer and leadership feedback to improve usability and flow.
• Common Component Adoption Across Consoles: Continued adoption of shared/common components to create a more consistent and unified experience across consoles.
• Hierarchical Agents, Phase 1: Introduced Phase 1 of the hierarchical agent orchestration layer, laying the foundation for more intelligent, coordinated experiences.
• Impact BU Uptake: Expanded onboarding and enablement to support Impact BU adoption and uptake of the platform and implementation capabilities.

• Tiered SKU Support for AI Native ITSM & CBS: Added support for Advanced and Prime SKUs across AI Native ITSM and CBS, including associated entitlement-driven flows and experiences.
• Product Hub UX Enhancements: Introduced targeted UX improvements to Product Hubs, including updates to the Resources section and guided content refinements based on leadership and customer feedback to create a more intuitive onboarding experience.
• LitJS Component Adoption in NAP: Enabled the ability to import and leverage LitJS components within NAP, helping establish a more scalable and consistent component framework.
• Admin Home Banner Updates: Updated Admin Home banners and messaging to better guide users through setup, actions, and key implementation milestones.
• Zero-Touch Implementation: Expanded Zero-Touch Implementation capabilities to deliver provisioned instances with entitled plugins and apps fully enabled, giving customers a value-ready environment from day one.
• Intelligent Configuration Update Tracking: Built on the April GA release with smarter tracking of configuration updates and changes, helping customers stay informed of new capabilities and recommended actions.
• Targeted UI Improvements: Delivered incremental UI enhancements across the experience, informed directly by customer and leadership feedback to improve usability and flow.
• Common Component Adoption Across Consoles: Continued adoption of shared/common components to create a more consistent and unified experience across consoles.
• Hierarchical Agents, Phase 1: Introduced Phase 1 of the hierarchical agent orchestration layer, laying the foundation for more intelligent, coordinated experiences.
• Impact BU Uptake: Expanded onboarding and enablement to support Impact BU adoption and uptake of the platform and implementation capabilities.

New:

• AI-powered response generation
• Smart Assessment Admin can activate/deactivate this capability at the template category level.
• Sources considered for response generation using AI
  • Previous Assessment Analysis: Automatically generates response suggestions by analyzing similar completed assessments either in classic or previous Smart Assessments
  • Document Intelligence Integration: Extracts relevant information from uploaded documents to suggest assessment responses
• Custom extension point to control the grounded sources
• Localization updates

New feature:

Entitlement creation error handling - This feature uses AI to suggest resolutions when users upload the standard Excel import template for entitlement creation.

New:

• The Intent to Action agentic workflow analyzes incoming emails to automatically identify the appropriate procurement intent and create corresponding procurement case records. Procurement teams can review the generated cases and take the necessary actions to efficiently manage procurement requests.
• The Savings Opportunity Discovery agentic workflow analyzes contracts, spend data, sourcing pipeline data, and supplier performance on a recurring schedule to automatically identify a prioritized list of savings opportunities. Category managers can review each opportunity in the Now Assist Panel and create a pipeline project or dismiss the opportunity directly from the panel.

Non Glide Cobalt Raven ACLs added.

New:

• Introduced Project Q&A conversational capability to ask questions about your project in natural language and get instant, AI-generated answers directly from the project workspace, without having to manually analyze project data.
• Use Portfolio insights to analyze planning items in a portfolio plan for execution delays, delayed starts, planned versus approved date misalignments, and projects at risk. View AI-generated root causes and recommended actions to mitigate identified risks and issues.

Changed:

• Project insights now includes scope creep as an additional topic and considers work notes at the task level as additional context, providing richer and more actionable project visibility.
• Updated formatting for in-app insight widgets to improve readability and presentation within the project workspace.
• Project Risk Detection now includes clickable links in AI Rationale for quick navigation to referenced project records. Refined Mitigation plan formatting for a better reading experience.

New:

• Ability to link emails to the created cases and the supplier records
• Ability to create assessment templates and error handling while cancelling created assessments

Migrated Generate Talking Points and Text to Growth Plan skills to the Mosaic AI framework.- Added Interview Scheduling enhancements.- GenAI evaluation framework: created GenAI MATs and auto-eval runs for Job Requisition Creation and Interview Scheduling agentic flows.

New support of 3p models and security fixes

Small models:Claude Haiku 4.5GPT_5 miniLarge Models:Claude Sonnet 4.6GPT_5.4

Mosaic migration done for couple of Skills

Product Release Content Generator

Renewal Insight Engine

Added skill categorization metadata for improved organization.

Enhanced application visibility by setting private flag to restrict access.

Resolved duplicate issue recommendations when multiple issues generated for a single question (PRB2015422).

Initial Release

• Use Now Assist to Vault to enhance your security posture autonomously by identifying, classifying, and protecting sensitive data in your custom applications.
• Surface sensitive data access by users automatically by leveraging Now Assist to configure, audit, and summarize your Access Observer logs.

New

• Admins can configure pronunciation dictionaries for voice assistants. Define custom pronunciations for domain-specific and company-specific terms using phonetic spelling or phoneme format. Entries are scoped per language and deployment and applied during voice interactions.
• AI voice assistants now support SIP call handling. Configure SIP transfer targets and methods, and expose SIP call data for integration with CCaaS providers.
• Voice deployments support multiple secondary languages with dynamic selection prompts. APIs return translated prompts, welcome messages, and tool execution text for all configured secondary languages. Language selection prompts are generated dynamically based on available languages and trigger phrases.
• Russian, Danish, and Turkish voice options are now available. Each language includes male and female voice options for different conversational contexts.
• New deployment attributes are available for advanced voice configuration. Configure noise suppression levels, orchestration and gateway modes, and selectively disable authentication flows for specific deployments.
• Filler message delay is now configurable per deployment. Set the wait time before filler words are spoken during processing-intensive operations to control conversational pacing.
• Voice assistants now support PSTN request processing. The system selects the appropriate request processor based on the active communication channel, enabling PSTN alongside existing providers.

Changed

• Voice analytics dashboard now organises performance metrics across four tabs Overview, Performance, Insights, and Assist consumption with new metrics for resolution rate, authentication performance, tool execution, conversation sentiment signals, and AI agent assist consumption.
• execute_tool and execute_tool_guest API responses now include execution plan and task identifiers. Use these identifiers to track and troubleshoot tool executions.
• Voice language configurations can now be deleted across application scopes. Only voice admins have permission to delete configurations from other scopes.
• Execution latency metrics are now captured for agent evaluation. The system records execution time data for voice agent plans to support performance analysis.

Fixed: Corrected language selection prompts for secondary languages to align with voice-based language selection behavior.

Removed

New:

The Security Exposure 360 agentic workflow brings AI-powered exposure analysis to USEM. Users can now ask questions in plain language and get answers grounded in their own ServiceNow data across all types of findings within USEM.

• New
• Changed
• Fixed
• Removed

Changed

Suggestion readers are now mapped automatically when Premium Chat is enabled for global search, a workspace, or a service portal, so suggested content surfaces without additional setup.

• Resolved an issue that prevented hybrid search from being enabled for some search application configurations.
• Suggested Utterances now carry over correctly when a portal is upgraded to Premium.

Made minor enhancements.

• Changed
  • Identify missing clauses in contract revisions with improved accuracy.
  • Support for upgraded versions of third party LLMs has been provided.
• Fixed
  • The activity stream now correctly displays the document name when an AI review is initiated, replacing the previous 'null' placeholder.
  • Email notifications to the contract administrator are now reliably sent when metadata extraction is completed.

• Changed
  • Updated default model configuration per ServiceNow platform guidance
  • Removed UI bounding box visualization feature to resolve node restart issues caused by memory contention during candidate mapping
• Fixed
  • Removed the legacy "enable_multimodal" key in the configuration
  • Extracted field values returning "empty" instead of "missing in document" when no value is found in the source document

No new features are added as part of this release.

Knowledge generation capabilities through Now Assist panel and in-product

Use the Standard Ticket Summary Fields related list to configure summary fields on the standard ticket page such that the page displays the Summarize button, enabling requesters to view Now Assist-generated overviews of recent ticket activity and details.

New

• Premium Chat available for portal, mobile, and Microsoft Teams. Options reflect configurations & AI Search status. Enable/disable per channel. Not for guest users; Virtual Agent notifications are not yet supported.
• Voice input toggle at the assistant level for premium chat. End users activate it in preferences.
• Unlimited asset promotion. Assets appear as pills in standard/enhanced chat, suggested topics in premium chat.
• Configure greeting and fallbacks for premium chat using OOB scripts, static text, or topics. Legacy configs pre-filled on upgrade.
• Mobile chat embed supports theming, analytics, and record context via URL parameters.
• Toggle contextual suggestions in Now Assist panel for autosuggest and homepage.
• Inactivity timeout max now 300 seconds.
• Banner alerts when premium chat setup in progress.
• 'All assets' view in assistant edit flow with empty state handling.
• Only active assets can be promoted; inactive assets are hidden from tables.

Changed

• Display experience config updated for portals/mobilereflects customer type, configuration and AI Search status. CSM portals restrict certain experiences.
• Premium chat logic revised for mobile/portal. Custom apps option removed.
• Tooltips updated for conversational catalog Itemsclarify parity gaps & form interactions.

Fixed

• Catalog auto-start with sources PRB2013012
• Skill selection recording PRB2006757
• Zero-result timeouts PRB2017263
• Voice setup mobile-only PRB1987951
• Enhanced chat end response PRB1994560
• Premium chat upgrade greeting/fallback PRB2028521
• CSM portal external user PRB2023644
• Duplicate Now Assist panel channels PRB2026739
• Asset tab refresh and add card PRB2029087
• Asset ordering & unsupported types PRB2028691
• Review tab details PRB2028587
• Chat menu items PRB2014288
• Provider unlinking PRB2026631
• Missing sources PRB2009215
• Voice tags PRB2019586
• Empty secondary language save PRB2024758
• Config attribute removal on channel save PRB2025395
• Genesys/3CLogic URLs PRB2023611
• Localization strings PRB2025243
• Amazon Connect PSTN block PRB2021363
• Genesys API Key switching PRB2021070
• Test button UI PRB2017785
• Tag addition PRB2016968
• AI Agent modal button PRB2017769
• Orphaned deployment configs PRB2011804
• External sources URL PRB2008332
• Learn more URL PRB2018834
• Search profile rename PRB2006849
• cloud_document_enabled default PRB2018953
• Zero genius results handling PRB2010786
• Branding menu items PRB2018544
• Channel save PRB2023061
• BT1 pipeline PRB2022683
• Databroker URI PRB2016144
• Branding menu items PRB2016228
• upgradeAISApp check PRB2016713
• Designer UI PRB1988383
• Zero query attribute PRB2029160

Removed

• 'sn_nowassist_va.enable_nap_aix_experience' property
• Custom PSTN
• Premium chat for custom apps mobile
• Inactive asset promotion

Added an agentic workflow for servicenow table record navigation

Support for Visual and Doc Qna

New:

Auto QA Breakdown Dashboard

Changed:

• UX improvements
• Functional improvements in Admin UI

Fixed:

UI Fixes related to data visualization

• Fixed an issue where the system got stuck in a loop during sensitive detection in topic execution.
• Restored missing code required for conversational testing.
• Fixed an issue where entity scoring ignored the label for reference choice fields.
• Corrected a mismatch where DateTime values were not displayed in the selected format.
• Fixed a looping issue where the Virtual Agent kept asking the same question despite receiving a user response.
• Improved the slot-filling prompt to better handle relative date inputs (e.g., tomorrow, next week).

Ability to evaluate agentic voice assistant and the agents within it.

• Evaluate standalone voice agents and voice assistants independently.
• Select out-of-the-box evaluation metrics.
• Use multiple forms of data input for evaluation, including file upload, synthetic generation, and manual entry.
• View evaluation results at both aggregate and individual scenario levels.

Ability to view the following latency numbers in evaluation run reports -

• Total session time: Measures the complete end-to-end task duration, including AI processing, tool usage, and user input delays.
• Agent processing time: Captures the time the AI agent spends working on the task, not counting user delays or waiting.

Changed

Updated metadata for OSM AI Advanced Application metadata has been updated to reflect current AI capabilities and configurations.

• Manage CRM records conversationally create, view, and update opportunities, contacts, leads, tasks, and accounts using natural language via a conversational interface. Use your MCP client to connect to the Sales CRM MCP server.
• Get AI-generated summaries of opportunity activity, bringing together emails, tasks, meetings, and product details without manual review.

New: Certified for Australia Patch 3

New:

• Agentic workflow to pro-actively identify escalations. The workflow evaluates various parameters like time remaining for SLA, urgency, priority, impact etc. to provides an escalation label (Critical, High, Medium) and reason for escalation.
• Skill to predict the next action including attaching relevant KBs, inking cases to Problem tickets, and updating form fields based on record context.
• Quick Wins skill that leverages an LLM to analyze task detailsincluding descriptions, short descriptions, journal fields, relevant KBs, and past resolution notesto determine if a task qualifies as a quick win
• Skill editing capabilities in NASK, providing a consistent editing experience across NAA and NASK
• Citations feature for summarization is now ready for uptake

Changed:

• Ability to filter on additional fields from record table (Asst group, Service, CI) for Analyze Task Trends
• Update Resolution notes activity filtering to include ai activity (include_ai_updates) config
• Updates to Citations to include related records
• Filter on additional params from record table for IWIS
• Summary of the generated work plan
• Auto classification of records GAF and config for Analyze task trends

Purchase Order Management - Foundation is the foundation-tier AI subscription for purchase order exception management on the ServiceNow AI Platform. Extending Now Assist for Purchase Order Management, this offering unlocks Platform Foundation features for enterprises that require the most advanced AI capabilities across their purchase order exception workflows. Purchase Order Management - Foundation adds access to Now Assist Platform Foundation, enabling organizations to take advantage of the latest platform-level AI innovations as they become available.

New:

Enhancement to the automatic conversion of supplier emails into PO Exception cases, now supporting additional languages and ERP purchase order IDs for improved PO identification.

• New
  • Implemented meta data changes to enable AI native support to all the skills and agents on privacy management product.
  • All the plugin dependency & role changes to enable AI Native capabilities are implemented as part of this application.
• Changed: The Now Assist issue summarization skill includes new configuration records offering descriptions and guides for configuration and prompt customization, enhancing the administrator and end-user experience without requiring upgrade actions.

Multilingual Enhancements

ICM Case Refinement Agent

New

Support Extended Fields

Mosaic initiative

New:

• Added new KG Query Decomposer Skill
• Added Execute compound query on KG and KB Subflow

Changed: Query range ACL's

• Consistent access control: All tables include standardized query range security ACLs. These ACLs ensure that authenticated users with appropriate read permissions can query records consistently across the platform.
• Seamless upgrade experience: New query ACL rules are installed automatically during upgrade, with no administrator action required. Automated upgrade scripts handle the transition, including detecting and processing previously customized ACLs to ensure existing processes continue without interruption.
• Post-upgrade review for customized ACLs: If the instance includes administrator-modified query range ACLs, review those records after upgrade to confirm they align with the intended access policy.

Changed

Updated the Query range ACL.

We do not have any changes in this app. Releasing this version as part of bundle

We do not have any changes in this app. Releasing this version as part of bundle

New

1. Image to Task plan template AU Agent- Allows users to convert a fulfilment flow image in a pdf, jpeg, png format to a task plan template.

Please refer to Advanced SKU

We do not have any changes in this app. Releasing this version as part of bundle

We do not have any changes in this app. Releasing this version as part of bundle

Changed

Updated the dependency version.

Changed

Updated the dependency version.

Changed

Updated the dependency version.

Connections tab in the Service Exchange Center

Create, view, request, and offboard provider and consumer connections from a single location in the Service Exchange Center. Search and filter connections without navigating across multiple screens.

Improved consumer registration and onboarding

Onboard consumers faster with a guided, step-by-step registration experience. Upgraded consumers are automatically redirected to this experience to receive clearer progress indicators during onboarding, and actionable messaging for failure and delay scenarios, minimizing onboarding friction and support dependency.

Improved FDS capabilities

• Improve your connection experience, by syncing Knowledge Base articles between provider and consumer instances.
• Reduce data inconsistencies by maintaining sys IDs for CMDB data and dependent relationships through transform maps.
• Ensure CI functionality is preserved on the destination instance by choosing to automatically create CI dependency relationships when relationship data is received from the source.
• Improved compliance through restricted data sync from non production instances to production instances for CMDB tables.

Journal Field Framework enhancements

• Increase flexibility in journal data synchronization between provider and consumer instances by mapping multiple source fields to a single target journal field.
• Configure journal fields such of type journal_input fields alongside journal type, ensuring all journal entries are preserved during synchronization without requiring custom scripting

Group-based persona assignments for Remote Catalog

Assign Remote Catalog personas to user groups so existing group-based access management practices extend to Remote Catalog, reducing administrative effort by managing access at the group level instead of individual users.

Connections tab in the Service Exchange Center

Create, view, request, and offboard provider and consumer connections from a single location in the Service Exchange Center. Search and filter connections without navigating across multiple screens.

Improved consumer registration and onboarding

Onboard consumers faster with a guided, step-by-step registration experience. Upgraded consumers are automatically redirected to this experience to receive clearer progress indicators during onboarding, and actionable messaging for failure and delay scenarios, minimizing onboarding friction and support dependency.

Improved FDS capabilities

• Improve your connection experience, by syncing Knowledge Base articles between provider and consumer instances.
• Reduce data inconsistencies by maintaining sys IDs for CMDB data and dependent relationships through transform maps.
• Ensure CI functionality is preserved on the destination instance by choosing to automatically create CI dependency relationships when relationship data is received from the source.
• Improved compliance through restricted data sync from non production instances to production instances for CMDB tables.

Journal Field Framework enhancements

• Increase flexibility in journal data synchronization between provider and consumer instances by mapping multiple source fields to a single target journal field.
• Configure journal fields such of type journal_input fields alongside journal type, ensuring all journal entries are preserved during synchronization without requiring custom scripting

Group-based persona assignments for Remote Catalog

Assign Remote Catalog personas to user groups so existing group-based access management practices extend to Remote Catalog, reducing administrative effort by managing access at the group level instead of individual users.

Connections tab in the Service Exchange Center

Create, view, request, and offboard provider and consumer connections from a single location in the Service Exchange Center. Search and filter connections without navigating across multiple screens.

Improved consumer registration and onboarding

Onboard consumers faster with a guided, step-by-step registration experience. Upgraded consumers are automatically redirected to this experience to receive clearer progress indicators during onboarding, and actionable messaging for failure and delay scenarios, minimizing onboarding friction and support dependency.

Improved FDS capabilities

• Improve your connection experience, by syncing Knowledge Base articles between provider and consumer instances.
• Reduce data inconsistencies by maintaining sys IDs for CMDB data and dependent relationships through transform maps.
• Ensure CI functionality is preserved on the destination instance by choosing to automatically create CI dependency relationships when relationship data is received from the source.
• Improved compliance through restricted data sync from non production instances to production instances for CMDB tables.

Journal Field Framework enhancements

• Increase flexibility in journal data synchronization between provider and consumer instances by mapping multiple source fields to a single target journal field.
• Configure journal fields such of type journal_input fields alongside journal type, ensuring all journal entries are preserved during synchronization without requiring custom scripting

Group-based persona assignments for Remote Catalog

Assign Remote Catalog personas to user groups so existing group-based access management practices extend to Remote Catalog, reducing administrative effort by managing access at the group level instead of individual users.

• New: Use Lens as a Service to support three auto-mapping services between Excel and ServiceNow tables: auto-map Excel column headers to ServiceNow table fields, auto-map Excel choice column values to ServiceNow choice field values, and auto-map Excel reference column values to ServiceNow reference field values.
• Changed: ServiceNow AI Lens now uses visual gradient indicators to distinguish AI-filled form fields from manually entered data.

Updated app/plugin dependencies

Updated app/plugin dependencies

New feature:

Entitlement creation error handling - This feature uses AI to suggest resolutions when users upload the standard Excel import template for entitlement creation.

New feature:

Entitlement creation error handling - This feature uses AI to suggest resolutions when users upload the standard Excel import template for entitlement creation.

Changed:: Updated application and plugin dependencies to ensure correct installation sequencing and compatibility.

Changed:: Updated application and plugin dependencies to ensure correct installation sequencing and compatibility.

New:

• Business users create templates to manage process so that all tasks are standardized across the organization. The template AI agent helps in creating templates from their current existing documents.
• Users can upload images or diagrams which contain a process flow/flow diagram. The Template AI agent reads these documents and analyzes the template and template items required.
• The Template ai agent creates a draft template along with template items based on the document and provides draft template details which are verified by the user.

We do not have any changes in this app. Releasing this version as part of bundle

We do not have any changes in this app. Releasing this version as part of bundle

We do not have any changes in this app. Releasing this version as part of bundle

New support of 3p models and security fixes

Small models:Claude Haiku 4.5GPT_5 miniLarge Models:Claude Sonnet 4.6GPT_5.4

New support of 3p models and security fixes

Small models:Claude Haiku 4.5GPT_5 miniLarge Models:Claude Sonnet 4.6GPT_5.4

New support of 3p models and security fixes

Small models:Claude Haiku 4.5GPT_5 miniLarge Models:Claude Sonnet 4.6GPT_5.4

New support of 3p models and security fixes

Small models:Claude Haiku 4.5GPT_5 miniLarge Models:Claude Sonnet 4.6GPT_5.4

We do not have any changes in this app. Releasing this version as part of bundle

We do not have any changes in this app. Releasing this version as part of bundle

We do not have any changes in this app. Releasing this version as part of bundle

Added skill categorization metadata for improved organization.

Resolved duplicate issue recommendations when multiple issues generated for a single question (PRB2015422).

Added skill categorization metadata for improved organization.

Resolved duplicate issue recommendations when multiple issues generated for a single question (PRB2015422).

Initial Release

New:

The Security Exposure 360 agentic workflow brings AI-powered exposure analysis to USEM. Users can now ask questions in plain language and get answers grounded in their own ServiceNow data across all types of findings within USEM.

No new features with this release. Updated the application to support bundling with other USEM applications.

New:

The Security Exposure 360 agentic workflow brings AI-powered exposure analysis to USEM. Users can now ask questions in plain language and get answers grounded in their own ServiceNow data across all types of findings within USEM.

Fixed minor system defects

Resolved an installation issue that prevented Voice Input from activating correctly on some instances.

The release notes of the related plugins can be viewed on the respective plugin's release notes.

The release notes of the related plugins can be viewed on the respective plugin's release notes.

New

• Added a new "Waiting for requester" activity state to support workflows requiring user follow-up or additional input.
• Added telemetry instrumentation for the in-product agentic experience to enable activity usage and engagement tracking.
• Introduced support for configurable visibility of the "Restart/Cancel Activity" option in the overflow menu to avoid conflicts with playbook activities.
• Completed foundational integration exploration between Suggested Actions AI Agent capabilities and the in-product experience.

Changed

• Updated headers across CSP, modal, and inline experiences from "AI Workflows" to "AI Activity" for improved terminology consistency.
• Renamed "Input required" to "Input needed" for clearer user messaging.
• Updated activity action text from "Cancel" to "Stop Activity."
• Refined follow-up action labels and completed several small UI and copy improvements to enhance overall usability and consistency.

Fixed

• Resolved workflow handling for requester wait-state scenarios to ensure activities properly reflect and manage pending user responses.
• Addressed overflow menu conflicts between activity controls and playbook activity options.

Removed outdated terminology and legacy labels associated with prior AI workflow naming and activity actions.

• Discover automation opportunities across customer service interaction channels chat (portal and messaging), voice, and email in addition to incidents and cases.
• Transcript-based analysis infers customer intent and resolution steps from conversational language, surfacing automation opportunities that aren't captured in structured fields.
• Channel-specific agent matching recommends chat, voice, or email agents based on each opportunity's characteristics.
• Channel-appropriate agent generation when no suitable existing agent is found.

Changed

• Mining scope expanded beyond cases and incidents to include the Interaction table.
• Matching pulls from a unified agent registry so chat and voice agents are evaluated alongside existing agents and tools, with no separate setup per channel.
• Clustering now incorporates interaction type, transcript content, short description, and referenced knowledge base articles when ranking opportunities by impact.

Fixed

• Resolved a defect causing opportunity ranking to display in reverse order.
• Resolved an issue where asset counts did not match across the UI.

New -

• AIOps AI Specialist (limited availavilblity) - The AIOps AI Specialist AI is ServiceNows purpose-built AI agent for IT Operations, designed to autonomously monitor, correlate, triage, and resolve infrastructure and application events at enterprise scale.
• Onboarding of the AIOps AI Specialist, including setting up the scope for the AI Agent, testing it and confgiure its task
• Remediation agent that works alongside the AIOps AI Specialist
• Proactive grouping recommendation using AI

Mosaic migration done for couple of Skills

Product Release Content Generator

Renewal Insight Engine

New:

AI Agent that investigates the missing attribute brought in through Pattern based discovery, finding affected CIs and links errors to standard error codes.

Updated to support the latest version of the dependent apps.

Fixed: Missing xml field that were causing data quality tests to fail

In this version, the following enhancements have been made:

• Security and access control management have been enhanced with simplified installation and automatic configuration handling.
• Stockroom locations no longer appear for software requests.
• The Orchestrator now displays human-readable output in work notes.
• Document extraction now works seamlessly with improved field configuration.

New

Added Splunk, AWS CloudWatch, ThousandEyes, SolarWinds, and Prometheus as supported investigation backends for the Analyze alert impact agentic workflow, expanding the alert insights available.

• Updated the Analyze alert impact agentic workflow to route all alert investigation through a single SRE Investigate AI Agent, delivering consistent behavior across all supported backends and simplifying the path for adding new vendor agents.
• Migrated New Relic and Datadog integrations to Model Context Protocol (MCP) agents, standardizing on vendor-maintained tool surfaces so the investigation workflows stay current as vendors ship updates.

Deprecated

Deprecated the Dynatrace Analysis AI Agent, superseded by the Dynatrace MCP Server Agent.

Connections tab in the Service Exchange Center

Create, view, request, and offboard provider and consumer connections from a single location in the Service Exchange Center. Search and filter connections without navigating across multiple screens.

Improved consumer registration and onboarding

Onboard consumers faster with a guided, step-by-step registration experience. Upgraded consumers are automatically redirected to this experience to receive clearer progress indicators during onboarding, and actionable messaging for failure and delay scenarios, minimizing onboarding friction and support dependency.

Improved FDS capabilities

• Improve your connection experience, by syncing Knowledge Base articles between provider and consumer instances.
• Reduce data inconsistencies by maintaining sys IDs for CMDB data and dependent relationships through transform maps.
• Ensure CI functionality is preserved on the destination instance by choosing to automatically create CI dependency relationships when relationship data is received from the source.
• Improved compliance through restricted data sync from non production instances to production instances for CMDB tables.

Journal Field Framework enhancements

• Increase flexibility in journal data synchronization between provider and consumer instances by mapping multiple source fields to a single target journal field.
• Configure journal fields such of type journal_input fields alongside journal type, ensuring all journal entries are preserved during synchronization without requiring custom scripting

Group-based persona assignments for Remote Catalog

Assign Remote Catalog personas to user groups so existing group-based access management practices extend to Remote Catalog, reducing administrative effort by managing access at the group level instead of individual users.

New

Added a Reliability Domain Agent that analyzes historical, post-incident data to identify service level objective (SLO) and synthetic monitoring gaps across configuration items (CIs) and services, then generates reliability tasks to address them.

Initial Release

• New
• Changed: Workplace Utilization QnA Agent can now be used by users with the sn_wsd_spcmgmt.space_planner role
• Fixed: The AI Agent was not always able to block the location after submitting an emergency maintenance request
• Removed

This is a "marker app" with no new functionality added; its purpose is to automatically trigger installation of other apps.

New:

• Refine actions You can now refine generated summaries directly within the dashboard using built-in actions such as Elaborate and Shorten, giving you more control over how the summary is presented.
• Feedback on generated content Added thumbs up / thumbs down controls to the summary component, making it easy to share feedback on AI-generated summaries.
• Dashboard Summary for UI Builder Workspaces Dashboard Summary is now available out of the box in UI Builder Workspaces.

Fixed:

• Consistent cache behavior across summary triggers Both the Refresh and Generate Summary buttons now apply the cache consistently, avoiding redundant AI calls.
• Dashboard Summary time zone aligned with user preference The Dashboard Summary now correctly applies the user's configured time zone rather than the system time zone, ensuring consistency with the data shown in dashboard widgets.

• Improved keyboard navigation in explorations:
  • Navigate explorations fully via keyboard, including text and action buttons.
  • Move up/down buttons have been added under the drag-and-drop icon.
• The data visualization sparkle icon is no longer shown for data sources in a non-supported secure scope.

• Record desktop actions more accurately by using the new AI-powered recording mode when creating desktop actions.
• Save time on manual setup by letting AI automatically insert anchors and generate screen context for each captured screen and add desktop action description after recording.
• Switch to AI-assisted recording by selecting the new Record with AI (recommended) check box that replaces the previous capture modes in the Create desktop action modal.
• Make desktop actions more flexible by configuring parameters for on-screen task desktop actions.
• Pass dynamic values at runtime by mapping parameters in the Map parameters section in AI Agent Studio.
• Control data visibility and security by using the Shared and Mark As Sensitive fields on the Desktop action parameter form.
• Get a quick guidance on how to effectively use the recorder with the recorder tips modal.

New features for June Release:

• Improved error handling and internationalization (i18n) support
• Added streaming capabilities for real-time data processing
• Resolved critical defects and stability issues
• Extended support for additional widget property types

• Addressed Brazil security directive
• Honoured document_match_count in the Uber RAG instead of default.
• Updated default chunking mode to STB(Small to big) instead of Fixed Size.

Changed

Deprecated searching and scraping functionality.

New

AIOps AI Specialist (L1 AI Worker)

AI-powered first responder that autonomously triages alerts, reduces noise, enriches context, and provides remediation recommendations across the alert lifecycle.

• Context-rich alert investigation Enhanced alert analysis using CMDB topology, incident history, and operational signals to provide deeper insights and faster decision-making.
• Knowledge-driven investigation Integrated knowledge sources, including KB articles into the investigation workflow to improve accuracy and deflect incidents
• Remediation suggestions from existing subflows AI-generated remediation recommendations leveraging existing alert rules and subflows to accelerate resolution and ensure consistency.
• SRE AI Specialist integration Integrates seamlessly with the SRE AI Specialist for deeper investigation, including observability data analysis and root cause hypotheses

Dedicated AIOps Manager Homepage

New homepage for AIOps Managers and Event Management Admins to onboard, configure, and manage AI Specialists from a centralized command center. _

_

New:

Proactive Grouping Recommendations AI-generated grouping recommendations help AIOps administrators identify alert correlation opportunities with greater speed and accuracy.

Changed:

Updated supported LLM versions to GPT-4o Mini and Claude Haiku 4.5, bringing the latest model improvements to the platform.

Fixed:

The Autonomous Operator was not automatically updating the description of HLA alerts. This has been corrected.

Fixed - Streamline ACLs

• Fixed ability to add to an exploration a list type data visualization.
• Fixed the ability to edit a data visualization in Visualization Designer.

• New
  • Upload files to Build Agent to provide context: images (PNG, JPEG, GIF, WEBP), documents (PDF, DOC, DOCX, XLS, XLSX), and text/code files (TS, JS, PY, JSON, MD, HTML, YAML, and more)
  • View update sets created by Build Agent from within the chat panel. Each checkpoint includes a button that opens the relevant update set in a new tab.
  • Connect Build Agent to external MCP servers in ServiceNow Studio. Previously, MCP server connectivity was only available in the ServiceNow IDE.
  • Validate user interface output during app creation with the UI validation tool in Build Agent, now available in ServiceNow Studio.
  • Use the Search retrieval tool to enable agents to fetch and present relevant information from configured data sources in response to user queries. Agents can surface knowledge articles, catalog items, and other indexed content directly within the agentic workflow, reducing the need for users to navigate to separate search interfaces.
• Changed: New MCP integrations in ServiceNow Studio: Added support for many new connectors including Atlassian Rovo, DocuSign, Figma, GitHub, Linear, Miro, Prisma Postgres, and Zoom (Chat, Docs, Revenue Accelerator, Whiteboard)

Restrict context based on requestor's data

• Support execution of Agent Chat, Sidebar Chat and Call Summarization to execute through Mosaic
• Restrict Agent Chat Summarization Card only to the agent invoking it
• Provide support to add default tables in Sidebar Chat Summarization configuration
• Uptake Cobalt Raven ACLs Brazil Directive on the table sys_generative_ai_predictions_log

New:

AI Framework that autonomously discovers schema, samples, and aggregates live ServiceNow tables, and presents a sourced answer without modifying any data.

Enhancements and defects fixed:

1. AI modified field indicator. AI Sparkle on Fields: AI-generated field indicators are now displayed consistently across both the platform UI and Workspace.
2. Complaint Case Intake Agent
   • Added the hyperlink to the complaint case number.
   • Prompt enhancements on location and products
   • Related Parties for Consumers.
3. Complaint Case Research Agent
   • Deleted default tasks after the creation of agent tasks.
   • Prompt enhancement on multi messages into one.
   • Search profile separation

Performance improvement.

New: Conversation Insights now supports customers with data sovereignty requirements. The application can be re-enabled for customers using local SKUs, allowing deployment with local conversation insights services.

• Changed: None
• Fixed: The assignment of the sn_aci.insights_read_write role to the conv-insights-svc-account user has been corrected to ensure proper handling of inheritance flag information.
• Removed: None

• New
  • General access for Premium Chat on Portals, Employee Slate, MS Teams, and Mobile
  • Voice Input for Premium Chat
  • Voice Assistant support on the 3rd party mobile app.
  • New metrics supported for the conversation auto evaluation feature
• Changed
  • Now Assist Panel Platform (default) assistant would be turned on by default for new customers
  • UX improvement on the Assistant edit flow
  • Inactive assistants can be tested in the Testing panel
  • Assets can be assigned to inactive assistants.
  • Setup topics without Gen AI skill will have visible and discoverable false, toggle will be disabled. For Setup topics that have Gen AI skills, admin can still toggle these flags.
• Fixed
  • Number of Promoted Assets is no longer restricted from the Assistant edit flow
  • Accessibility color contrast

Changed: Maintenance release.

Changed:

• Added voice-driven case status retrieval and updates, enabling customers to check open case statuses and submit updates through guided voice interactions across Genesys, NICE, Five9, 3CLogic, and Amazon Connect, reducing live agent dependency.
• Enhanced Provide Customer 360 Insights with Enterprise Graph and AI Agent deep research for richer, more contextual query results.

• Agent renamed to "Vision AI Agent"
• Requests now routed via the MultiModal Service (MMS) to enable more consistent and scalable multimodal processing (up to 500MB video files)
• Improved response quality for Video Workflow use cases
• Supported video formats: MP4, MOV, WEBM, AVI, MKV, and WMV

Fixed:

Updated the version for ServiceNow Docs Connector to fix dependency failures

This release provides the following enhancements and improvements.

• Enhanced accessibility support Performance improvements Support for L10n Enhanced telemetry support for tracking user behavior Enhanced the home page layout with auto-adjust flexibility based on the widgets count Enhanced Admin configuration to support Tasks and requests (fka Inbox), To-dos task type configurations
• Enhanced AI Preferences to support Smart Prioritisation and AI-Generated Summarisation
• Enhanced the option to load AI summary automatically or on demand on AI summary details page

• New: Improvements to the AI Agent to support NextWave compatible experience.
• Fixed
  • Agent prompt quality improvements.
  • Hallucination issues to improve AI Agent performance.
• Removed: Implicit dependency on the Webcrawler external content connector. The Webcrawler connector now needs to be installed explicitly as a soft dependency.

• AI-powered mobile form auto-fill for field technicians.: Field technicians can use mobile camera to capture images and auto-populate form fields via AI image recognition, supporting smart assessments, input forms, and scripted questionnaires.
• Consistent AI visual indicators across applications.: AI features now display unified gradient styles and subtle animations to indicate activity in Workspace, UI16, and mobile experiences, meeting accessibility standards.
• Parts Debrief automation in NAP and NAVA.: Agents can automate parts debrief workflows, handling empty work notes, invalid part usage, quantity normalization, and error scenarios.
• Expanded AI summarization and knowledge workflows.: Users can trigger work order/sidebar chat summarization and knowledge article generation across platforms with role-based workflows and answer validation.
• Create work order agent with enhanced persona coverage.: System supports creating work orders from text/images in NAP/NAVA with persona validation, blocking invalid content.
• FSM AI Mandate features available on mobile and workspace.: FSM AI capabilities (IT automation, NAVA integration, smart assessments) are now mobile/workspace accessible with role-based access.
• Telemetry for AI feature usage.: Admins can track AI lens feature usage for FSM use cases.
• Skill refinement for FSM digital agents.: FSM agents enhanced for improved automation and workflow handling.
• Changed
  • Broadened test and validation coverage for FSM AI use cases.: Testing expanded across mobile/agent/NextWave experiences to ensure certification and release standards.
  • Enhanced automation for AI SKU and agent studio test cases.: Remaining automation scenarios completed with monitoring on latest NAP version.
  • Improved parts debrief workflow handling.: Better management of edge cases like unrecognized assets and invalid task states.
  • Refined questionnaire answer validation in summarization workflows.: Validation improved across NAP/NAVA/UI16/Workspace for task summarization.
  • Performance testing for Q1 deliveries.: Testing conducted to ensure system stability and responsiveness.
• Fixed
• Removed

Initial release

New: Guardian prompt injection detection is being enabled as always-on and in blocking mode by default for Skills. Added granular admin controls to enable Guardian's prompt injection detection independently for skills, agents, or both

Changed: Increased Gen AI execution log (sys_generative_ai_log) retention to six months.

New:

Backend support for HR Knowledge Gaps. This capability will be taken up by the Knowledge Center team in a future release. No customer-facing updates in this release.

Changed:

Nothing changed in this release.

Fixed:

Various minor defects fixed.

Removed:

Nothing removed in this release.

Resolved defects affecting content and agent tool functionality

New: Added new skills for Interview health tracker.

Fixed: AI Agent records view was affected due to an AI Native ACL issue

New:

Updates to support AI Agent Advisor.

Changed:

Nothing changed in this release.

Fixed:

Nothing fixed in this release.

Removed:

Nothing removed in this release.

1. The L1 Service Desk AI Specialist now detects knowledge gaps and automatically creates a Knowledge Feedback Task when KB search returns no results above the configured relevance threshold, enabling managers to create missing articles.
2. Knowledge feedback tasks are now deduplicated the system checks for existing open tasks with matching category and subcategory within a lookback window before creating a new one, preventing task noise for gaps like password resets.
3. The Change Quality Agent now persists quality scores per change record including timestamps and contributing factors available for dashboards, audits, and risk-detection workflows.
4. A new KB article usage section on the AI Agent Studio Performance dashboard shows which articles the AI Specialist used during autonomous resolution.
5. The AI Specialist now persists the Skill applied per incident evaluation onto the record. A new Skill filter on the AIS Performance Dashboard lets managers segment performance by capability area.
6. The Insights and Opportunities for Incident dashboard in Service Operations Workspace aggregates incidents into trend categories with AI-generated summaries, SLA performance, sentiment, channel adoption, and geographic distribution, giving incident managers a consolidated view of incident patterns with drill-down views into individual records.
7. IT administrators can now configure Change Management through an AI-native conversational agent in the Product Console, guided through approvals, risk scoring, and workflows via natural language.
8. The ZTS AI L1 Service Desk Specialist now supports real-time conversational engagement via Microsoft Teams or Slack, replacing the asynchronous activity stream with a consent-driven experience.
9. A new agentic solution for SharePoint access issues autonomously handles L1 scenarios including access requests, file failures, and permission mismatches, reducing resolution time from hours to minutes.

1. The AI Specialist Performance Dashboard now includes revised productivity metrics measuring incidents attempted against total count for a more accurate resolution baseline.
2. Targeted enhancements to the L1 Service Desk AI Specialist improve reliability and consistency based on quality feedback from recent releases.
3. The Create Incident AI Agent has been migrated to a Hierarchical Agent model, resolving hallucination and rendering issues on NextWave and removing the VA dependency.
4. The Create Incident AI Agent has been updated for full NextWave off-glide orchestrator compatibility, resolving functional deviation issues from GA readiness testing.

• Fixed- None
• Removed - None

• New: n/a
• Changed: Security related enhancements on existing AI Voice agents
• Fixed: n/a
• Removed: n/a

• Service Mapping supports the Model Context Protocol (MCP), providing a conversational interface for live application service data. Service Mapping admins can connect Claude Desktop to a ServiceNow instance and query service topology, server-to-service relationships, related CIs, and unmapped configuration items.
• Service Mapping Workspace widgets1. Business app linked to App services - Indicating the count of business applications that were linked to application services by the Business App Mapping AI Agent.2. Maps created by Now Assist - Indicating the count of the service maps created by Service Mapping AI Agent.
• Changed
  • The AI specialists for Service Mapping are renamed as AI Agents for Service Mapping.
  • View the service map created by Service Mapping AI Agent through the AI Activity list.

The Knowledge Center centralized hub for Knowledge Admin and Manager:

Changes:

Added changes for better Typo handling for Table Labels.

Added internal usage tracking enhancements.

New

• Discover and execute Ansible playbooks from LEAP automation opportunities to promote faster incident resolution
• Track missed automation opportunities in LEAP to identify gaps in automation coverage
• Auto-generate resolution steps for all critical and high severity automation opportunities post initial run
• Improved resolution step relevance as irrelevant steps filtered out by group problem description
• LEAP available with Now Assist for ITSM

Changed

• Product renamed from AIOps LEAP to LEAP (Learning-Enhanced Automation Platform)
• AI-enhanced and structured knowledge base articles published to the Knowledge Center
• Fixed
  • Automation Opportunities page load time reduced from 3+ seconds
  • Value Dashboard page load time reduced from 4+ seconds
  • Resolved stale resolution steps reappearing in record metric table after SOW Flow updates
  • Fixed sharp-edged AI Gradient button for Regenerate in AO Details to match default theme
  • Fixed UI issues on resolution steps filter in LEAP Settings page
  • Fixed missing help content on LEAP Installer Skill Activation page
  • Reverted unintended default UI filter that auto-generated resolution steps
  • Fixed Maintenance ACL Roles being overridable by Admins in LEAP Automation Playbooks

AI Filter Assist allows users to define conditions using natural language in an intuitive, dialog-driven interface. This feature helps reduce the cognitive load of building conditions using the traditional condition builder by allowing users to type or speak a prompt that will then create the query needed to filter, sort, or group data.

This release we fixed two defects in AI Filter Assister where an ACL had a typo in it's script and "Add to existing filters" mode was selectable even though it should be disabled.

• New: Delivery note upload for quantity disputes: Customers can upload delivery notes during case creation, enabling the invoice dispute intake assistant AI agent to validate and resolve quantity disputes without human intervention.
• Changed: Order line quantity validation for disputes: The invoice dispute support assistant AI agent uses order line data when sold product records are unavailable such as new orders, non-serialized products, or services. It auto-approves valid disputes and escalates only when data is unclear.

• New: Automated email notifications for order cases: Send emails when the AI agent opens an order case from a Business Portal chat and when the case is resolved. Closure emails include resolution and quote details, if generated. Voice-created cases dont trigger emails.Scripted extension point for quote thresholds: A scripted extension point to integrate the manage order operations AI agent with your inventory, ERP, and quote systems for quote threshold evaluation. The AI agent generates a quote when a quantity-change request exceeds the configured price threshold.
• Changed
• Order exception support for quantity and shipping location: Customers can request quantity and shipping location updates, along with expedite requests via chat and voice channels.
• Voice assistant intake for order exceptions: Customers can submit expedite, quantity, and shipping location requests via voice. Requests are captured and converted into order cases for resolution in CSM/FSM Configurable Workspace.
• Unified extension point for feasibility checks: A single scripted extension point (sn_ord_ops_aias.orderExceptionCheckEP) validates delivery, quantity, and shipping feasibility, replacing the earlier expedite-only ATP check.
• Auto-populated account and contact on interactions: Account and Contact fields are populated during chat handoff, giving agents immediate customer context and improving summarization accuracy.
• AI-generated summaries for agent handoff: Agents receive a concise AI-generated summary instead of full chat history, improving context and productivity.

New:

Access Strategic Portfolio Management data and Now Assist AI skills as MCP tools, enabling LLM agents to query and reason about goals, portfolio plans, and projects. The following tools are available in this release: Tool Description get_goals Retrieves goals and objectives. generate_goal_insights Generates AI-powered insights for goals and targets. get_portfolio_plans Retrieves portfolio plans. generate_portfolio_insights Generates portfolio insights, including at-risk projects, delayed starts and ends, and dependencies. get_projects Retrieves projects. generate_project_insights Detects project risks, analyzes status trajectory, and provides recommendations. identify_project_risks Detects AI-identified RIDAC risks and saves them to the risk table as AI drafts. get_ai_status_report Generates a RAG (Red, Amber, Green) status report across resources, cost, schedule, and scope.

Removed : api-version query parameter from Create Response OEM action

New

Enabled the mobile card generation capability to be executed Off-Glide to improve scalability and performance.

Changed

Minor API changes for another app support internally

Performance improvements implemented, and third-party IDPs now allowed to be used.

New:

• Multi-intent identification support
• Missing inputs extraction
• email branding/template support for email generation

Instance and system administrators can now archive custom skills directly from the Now Assist Skills page. Archiving is not supported for ServiceNow out-of-box (OOB) skills. Archived skills can be restored at any time.

View Detail button is now available alongside the Activate button for OOB skills on the Now Assist Skills page. The Skill detail view provides administrators and users with key information about a skill including its benefits, prerequisites, recommended model provider, and related skills enabling informed activation decisions.

Prompt injection protection can now be configured at both the instance level and the product level (CSM, ITSM, and others), giving administrators more granular control over enforcement across workflows.

Administrators can now launch a guided tour walkthrough directly from the " Manage Model Versions" section within Settings section in now assist admin. The guided tour provides a step-by-step introduction to the model version management workflow, helping admins understand version lifecycle, deprecation alerts, and available actions.

Approval checklist generator skill accepts additional context parameter that is used to resolve checklists unknown items.

• New: AI Agent Extensibility: Customers can now add / remove AI agents or tools from OOTB agents while remaining on the upgrade path.
• Changed: Orchestrator defaults to GPT-5.4: The orchestrator now uses GPT-5.4 as its default model when Azure OpenAI is the selected LLM.
• Fixed
  • Agent Execution
    • Published version instructions not being applied to agents at runtime.
    • Execution plan fetch timing out on instances with large numbers of execution plan records.
    • Agentic workflow stuck at orchestrator step intermittently.
    • VA transaction timeout and conversation stuck on "Thinking."
  • Agent-to-Agent (A2A)
    • Async handling of working state from secondary agent causing primary flow issues.
    • Raw rich control output in agent response causing conversation to fault and close abruptly.
    • input-required response sending empty message/parts block.
  • Multilingual
    • Tool names, agent names, and "Processing new information" untranslated in AI steps.
    • Words not translated when user language differs from system language (intermittent).

• Browser tabs that open during goal execution in adaptive desktop actions remain open after the goal completes. Use the keep_tab_open system property to turn this behavior on or off. The property is turned on by default.
• Adaptive desktop actions is enhanced to improve execution efficiency.

• Analytics dashboards pertaining to AI agents, AI assistants, and skills are consolidated inside Now Assist Center > Monitor.
• Executions dashboard in Now Assist Center: Deep dive into individual executions of assistants and agents to improve performance.
• Business value dashboard in Now Assist Center: Analyze the impact of AI assets on time and cost savings and set formula for calculating value per AI asset (skills, AI agents, agentic workflows)

• AI asset inventory. A new Asset Inventory page provides a single, consolidated view of every AI asset deployed across the instance including AI agents, agentic workflows, skills, subflows, actions, data assets, virtual assistants, topics, catalog items, and knowledge graphs. Each asset type surfaces active/inactive status, and the list supports column personalization so administrators can tailor the view to their needs. Assets from AI Agent Studio, Data Kit, and the platform catalog are unified in a single browsable surface.
• AI readiness assessment. Administrators can now validate their instance's readiness for AI adoption directly within Now Assist Center. The new AI Readiness page supports on-demand and scheduled assessment execution across both Now Assist and Agentic AI assessment types. Results are surfaced in a dedicated Assessment Details page with historical tracking, enabling teams to identify configuration gaps, measure readiness over time, and accelerate the path to their first production AI solution.
• AI Agent Advisor enhancements. The AI Agent Advisor experience has been refined based on feedback from the May release. Updates include an improved automation opportunities list page with better filtering and layout, redesigned Agent Advisor cards, a refreshed home page CTA experience that surfaces opportunities at a glance, a new Agent Advisor settings page, and stack metrics page improvements. These changes reduce friction in the discovery-to-deployment journey and improve the quality of agent recommendations surfaced to administrators.
• Expanded OOTB skill configuration. The conversational administration experience now surfaces additional out-of-the-box skill configuration paths, including skill availability configuration and display experience settings extending the reach of guided, in-product setup to more of the Now Assist skill catalog.

Defects fixes and off glide LLM execution migration

Increased sample data input limit for SDG requests Users (Data Kit Admin) can now select up to 50 seed records from a data collection or table when submitting a Synthetic Data Generation request, up from the previous default of 3. The maximum sample size is also configurable by NADK Admins via sys_property.

Plausibility Metrics for Synthetic Data Introduced plausibility scoring as a new data quality insight layer within the Generated Data view (Data Kit > Home > Synthetic Data tab > Open Generated Data > Data Insight > Plausibility). Validates that generated field values are realistic, internally consistent, and free of garbage data.

OOB data sets

New:

• Leveraging Intent to Action Agentic Workflow for emails to be parsed and categorized as the correct intent to create a invoice inquiry case record.
• Enhanced workflow to enable requesters to accept/reject the case resolution.

Changed: Updated versions of dependent apps.

Now Assist for code generation features now inherit skill-level ACLs configured in Now Assist Admin, ensuring restricted features are only available to authorized users.

New: AI-powered import of Tasks and Stories into CWM Boards from spreadsheets, documents, and images using Now Assist.

Changed: No new features, enhancements or fixes. Since there are updates to the AI agents, the version of this application must also be updated.

Fixed: Security fixes

Existing feature: with latest third party model versions

• New
  • Build Agent
    • Upload files to Build Agent to provide context: images (PNG, JPEG, GIF, WEBP), documents (PDF, DOC, DOCX, XLS, XLSX), and text/code files (TS, JS, PY, JSON, MD, HTML, YAML, and more)
    • View update sets created by Build Agent from within the chat panel. Each checkpoint includes a button that opens the relevant update set in a new tab
    • Connect_Build Agent_to external MCP servers in_ServiceNow Studio. Previously, MCP server connectivity was only available in the_ServiceNow IDE.
    • Validate user interface output during app creation with the UI validation tool in_Build Agent, now available in_ServiceNow Studio.
    • Use the Search retrieval tool to enable agents to fetch and present relevant information from configured data sources in response to user queries. Agents can surface knowledge articles, catalog items, and other indexed content directly within the agentic workflow, reducing the need for users to navigate to separate search interfaces.
  • Catalog Item Generation: Migrated Now Assist for Catalog Generation skill group (Create Catalog Item and Refine Skill) to Mosaic (off-Glide GenAI Controller) by enabling offglide routing via OneExtend
  • Playbook Generation: Playbooks now automatically generate a concise, AI-powered summary of a Playbook's purpose, structure, and logic, helping new process owners quickly understand existing workflows without manual documentation.
• Changed: Build Agent <ul><li>New MCP integrations in ServiceNow Studio: Added support for 12 connectors including Atlassian Rovo, DocuSign, Figma, GitHub, Linear, Miro, Prisma Postgres, and Zoom (Chat, Docs, Revenue Accelerator, Whiteboard)</li></ul>

Enhanced case insights card with customer details

Enabled sentiment analysis on email interactions.

Changed

Engagement tool for DEX agents to obtain employee approval before auto-triggering device actions. This change integrates with ZTS so will not be published until ZTS is GA.

Updated to support the latest version of the dependent apps.

Initial Release

• AI-powered mobile form auto-fill for field technicians.: Field technicians can use mobile camera to capture images and auto-populate form fields via AI image recognition, supporting smart assessments, input forms, and scripted questionnaires.
• Consistent AI visual indicators across applications.: AI features now display unified gradient styles and subtle animations to indicate activity in Workspace, UI16, and mobile experiences, meeting accessibility standards.
• Parts Debrief automation in NAP and NAVA.: Agents can automate parts debrief workflows, handling empty work notes, invalid part usage, quantity normalization, and error scenarios.
• Expanded AI summarization and knowledge workflows.: Users can trigger work order/sidebar chat summarization and knowledge article generation across platforms with role-based workflows and answer validation.
• Create work order agent with enhanced persona coverage.: System supports creating work orders from text/images in NAP/NAVA with persona validation, blocking invalid content.
• FSM AI Mandate features available on mobile and workspace.: FSM AI capabilities (IT automation, NAVA integration, smart assessments) are now mobile/workspace accessible with role-based access.
• Telemetry for AI feature usage.: Admins can track AI lens feature usage for FSM use cases.
• Skill refinement for FSM digital agents.: FSM agents enhanced for improved automation and workflow handling.
• Changed
  • Broadened test and validation coverage for FSM AI use cases.: Testing expanded across mobile/agent/NextWave experiences to certify features.
  • Enhanced automation for AI SKU and agent studio test cases.: Remaining automation scenarios completed with monitoring on latest NAP version.
  • Improved parts debrief workflow handling.: Better management of edge cases like unrecognized assets and invalid task states.
  • Refined questionnaire answer validation in summarization workflows.: Validation improved across NAP/NAVA/UI16/Workspace for task summarization.
  • Performance testing for Q1 deliveries.: Testing conducted to ensure system stability and responsiveness.
  • Fixed
  • Removed

Minor defect fixes

This release delivers UI modernization and quality improvements. The Now Assist for HAM interface has been upgraded to the latest UI framework, with improved parsing and rendering of Asset Summary responses. Accessibility improvements have also been applied.

Improvements:

1. Currency fields extracted from purchase order attachments now display correctly.
2. A missing field configuration in the purchase order attachment flow has been added.
3. A component reliability issue in the record review workflow has been fixed.

New:: Contextual Action Planner sidebar in the Health and Safety Workspace: Create actions manually or from AI suggestions, and review them in context across incidents, observations, investigations, cases, audits, work permits, risk assessments, meetings, etc.

Consistent model support and on prem support

Security guardrails and Defect fixes related to content and metrics

Gen AI Trace logs improvement

Code Fix Agent: Enhanced Error Logging

When the Code Fix Agent runs into a problem, you will now see a plain-language explanation of what went wrong and what to do next right on the remediation record. Select the links navigate directly to the relevant error record, so there's no need to hunt through system logs.

New:

• AIOps AI Specialist Autonomous Alert Management (Limited Availability) The AIOps AI Specialist is ServiceNow's purpose-built AI agent for IT Operations. It autonomously monitors, triages, and resolves infrastructure and application alerts, executing the full NOC workflow without requiring human intervention on every event. In this release, the AI Specialist can:
  • Automatically pick up alerts based on configurable scope and filter rules
  • Run alert analysis, investigation, and impact assessment autonomously
  • Collaborate with the SRE Agent for deep-level investigation using A2A integration with observability tools and hyperscalers
  • Hand off to a remediation agent to execute subflow-based fixes
  • Close alerts determined to be insignificant noise
  • Be triggered manually by assigning an alert directly to the AI Worker user
• Onboarding Experience for the AIOps AI Specialist A guided setup flow helps NOC team managers configure and activate the AI Specialist quickly and consistently.
• Remediation Agent A new remediation agent works alongside the AIOps AI Specialist to automatically identify and execute the appropriate remediation subflow for an alert.
• Proactive Grouping Recommendations Alert Assist now proactively analyzes incoming alerts and recommends grouping candidates before a NOC operator opens a case. The engine uses intelligent field selection to identify the most relevant alert attributes, supports all LLM types, and handles high-cardinality and verbose alert data gracefully. Recommendations are refreshed every 5 minutes via a scheduled job.
• AI Resolution Attribution When the AIOps AI Specialist closes an alert, the system records it as "Closed by AI," capturing which alerts were fully handled autonomously. This enables NOC teams to measure automation rate and track AI-driven closure trends over time.
• Text to Configuration Agent Available out of the box, the Text to Configuration Agent allows operators to configure AIOps settings using natural language no form-filling required.

Changed:

• The AI Worker alert management rule now supports coexistence with autonomous mode.
• The auto-close path for insignificant noise alerts has been updated for more reliable handling.
• Autonomous Styled Output now generates a remediation anchor to support downstream remediation workflows.
• LLM Model Upgrades Alert Assist skills have been updated to use gpt-4o-mini (previously gpt-small) and Claude Haiku 4.5 (previously claude-small), improving output quality and aligning with current model availability.

Fixed:

• The AI Investigation Report displayed an infinite loading spinner on timeout.
• The Regenerate action in the AI Insights card was non-deterministic. The Express List also showed incorrect insight data when the first analysis run failed.
• The Autonomous Operator was not updating descriptions on individual HLA alerts.
• Alert Assist was logging unnecessary errors when no related changes passed the relevance threshold.
• Alert Assist was logging null pointer exceptions during skill availability checks.

1. The L1 Service Desk AI Specialist now detects knowledge gaps and automatically creates a Knowledge Feedback Task when KB search returns no results above the configured relevance threshold, enabling managers to create missing articles.
2. Knowledge feedback tasks are now deduplicated the system checks for existing open tasks with matching category and subcategory within a lookback window before creating a new one, preventing task noise for gaps like password resets.
3. The Change Quality Agent now persists quality scores per change record including timestamps and contributing factors available for dashboards, audits, and risk-detection workflows.
4. A new KB article usage section on the AI Agent Studio Performance dashboard shows which articles the AI Specialist used during autonomous resolution.
5. The AI Specialist now persists the Skill applied per incident evaluation onto the record. A new Skill filter on the AIS Performance Dashboard lets managers segment performance by capability area.
6. The Insights and Opportunities for Incident dashboard in Service Operations Workspace aggregates incidents into trend categories with AI-generated summaries, SLA performance, sentiment, channel adoption, and geographic distribution, giving incident managers a consolidated view of incident patterns with drill-down views into individual records.
7. IT administrators can now configure Change Management through an AI-native conversational agent in the Product Console, guided through approvals, risk scoring, and workflows via natural language.
8. The ZTS AI L1 Service Desk Specialist now supports real-time conversational engagement via Microsoft Teams or Slack, replacing the asynchronous activity stream with a consent-driven experience.
9. A new agentic solution for SharePoint access issues autonomously handles L1 scenarios including access requests, file failures, and permission mismatches, reducing resolution time from hours to minutes.

1. The AI Specialist Performance Dashboard now includes revised productivity metrics measuring incidents attempted against total count for a more accurate resolution baseline.
2. Targeted enhancements to the L1 Service Desk AI Specialist improve reliability and consistency based on quality feedback from recent releases.
3. The Create Incident AI Agent has been migrated to a Hierarchical Agent model, resolving hallucination and rendering issues on NextWave and removing the VA dependency.
4. The Create Incident AI Agent has been updated for full NextWave off-glide orchestrator compatibility, resolving functional deviation issues from GA readiness testing.\

1. Platform skills not appearing in the Now Assist Skills admin panel (even with ignoreFulfillerSubscriptionCheck enabled) have been resolved.
2. Display and interaction issues in the Recommended Actions panel have been resolved: misaligned loading indicator, non-functional search icon, footer-obscured filters, and incorrect cursor on non-clickable content.

• Changed: Support for upgraded versions of third party LLMs has been provided.
• Fixed: Legal request and matter summarization now delivers improved accuracy across all supported models

Automate the end-to-end creation, validation, and resolution of order exception cases on the Business Portal for Expedite, Additional Quantity, and Ship to Different Location requests.- Capture customer inputs in the Virtual Agent, extract intent and values inline, validate them against inventory, availability, location, and unit-of-measure rules, and confirm with the customer before acting.- Run a pricing threshold check and, when required, generate a quote for customer approval before resolving the case and updating case lines to Resolved.- Hand off seamlessly to a live agent at any failure or rejection point, with full conversation context, case number, and a clickable case link in NAVA preserved.

Updated App/Plugin dependencies

New:

• Total session time: Measures the complete end-to-end task duration, including AI processing, tool usage, and user input delays.
• Agent processing time: Captures the time the AI agent spends working on the task, not counting user delays or waiting.
• Customers will be able to view latency numbers in evaluation run reports (Agentic Evaluation). Two latency scores will be available:
  • Agent Instruction Optimization - added support for the latest third-party (3P) models, Hierarchical Agents, and Voice Agents.
  • 'Suggested Fixes' in Agentic Evaluation - added support for Hierarchical Agents.
  • Added Skill Group capability to update LLM providers across all Agentic Judges simultaneously through Now Assist Admin.
• Changed
  • Modified Overall Task Completeness evaluation for improved evaluation accuracy.
  • Agent Instruction Optimization - Enhanced gatekeeping mechanisms for AIO Log Analyzer, Agent Instruction Generator, Tool Description Generator

NOTE: These features and additions will be usable within Now Assist Skill Kit > Agentic Evaluation

Multilingual Enhancements

ICM Case Narrative Ai Agent Support

New:

The automatic conversion of supplier emails into PO exceptions workflow now supports:

• Additional languages, enabling better recognition of non-English supplier communications
• ERP purchase order IDs, allowing precise matching with purchase order data from enterprise resource planning systems

• Fallout creation AI Agent - This will be automatically generate fallouts in the correct category depending on the Notes/Comments on the Task.
• Document to Task plan template AI Agent - This Agent will be used to create a Task plan template using an document ( pdg, png,jpeg) which contains the fulfilment view of the Specification.

New dependant plugin - Opportunity Management AI Features added.

• Submitting feedback on AI-generated recommended actions fails for non-admin users (e.g., analysts).
• Correlation Insights - No output for certain CIs.
• Generate Insights action visible even when Now Assist for Security Incident Response was not installed.

Changed

• Summarization skill enhancement to honor chronological order of events and to give more weightage to recent activity.
• AI Gradient styling for Correlation Insights and Summarization.

Connections tab in the Service Exchange Center

Create, view, request, and offboard provider and consumer connections from a single location in the Service Exchange Center. Search and filter connections without navigating across multiple screens.

Improved consumer registration and onboarding

Onboard consumers faster with a guided, step-by-step registration experience. Upgraded consumers are automatically redirected to this experience to receive clearer progress indicators during onboarding, and actionable messaging for failure and delay scenarios, minimizing onboarding friction and support dependency.

Improved FDS capabilities

• Improve your connection experience, by syncing Knowledge Base articles between provider and consumer instances.
• Reduce data inconsistencies by maintaining sys IDs for CMDB data and dependent relationships through transform maps.
• Ensure CI functionality is preserved on the destination instance by choosing to automatically create CI dependency relationships when relationship data is received from the source.
• Improved compliance through restricted data sync from non production instances to production instances for CMDB tables.

Journal Field Framework enhancements

• Increase flexibility in journal data synchronization between provider and consumer instances by mapping multiple source fields to a single target journal field.
• Configure journal fields such of type journal_input fields alongside journal type, ensuring all journal entries are preserved during synchronization without requiring custom scripting

Group-based persona assignments for Remote Catalog

Assign Remote Catalog personas to user groups so existing group-based access management practices extend to Remote Catalog, reducing administrative effort by managing access at the group level instead of individual users.

• Tiered SKU Support for AI Native ITSM & CBS: Added support for Advanced and Prime SKUs across AI Native ITSM and CBS, including associated entitlement-driven flows and experiences.
• Product Hub UX Enhancements: Introduced targeted UX improvements to Product Hubs, including updates to the Resources section and guided content refinements based on leadership and customer feedback to create a more intuitive onboarding experience.
• LitJS Component Adoption in NAP:Enabled the ability to import and leverage LitJS components within NAP, helping establish a more scalable and consistent component framework.
• Admin Home Banner Updates: Updated Admin Home banners and messaging to better guide users through setup, actions, and key implementation milestones.
• Zero-Touch Implementation: Expanded Zero-Touch Implementation capabilities to deliver provisioned instances with entitled plugins and apps fully enabled, giving customers a value-ready environment from day one.
• Intelligent Configuration Update Tracking: Built on the April GA release with smarter tracking of configuration updates and changes, helping customers stay informed of new capabilities and recommended actions.
• Targeted UI Improvements: Delivered incremental UI enhancements across the experience, informed directly by customer and leadership feedback to improve usability and flow.
• Common Component Adoption Across Consoles: Continued adoption of shared/common components to create a more consistent and unified experience across consoles.
• Hierarchical Agents, Phase 1: Introduced Phase 1 of the hierarchical agent orchestration layer, laying the foundation for more intelligent, coordinated experiences.
• Impact BU Uptake: Expanded onboarding and enablement to support Impact BU adoption and uptake of the platform and implementation capabilities.

• Tiered SKU Support for AI Native ITSM & CBS: Added support for Advanced and Prime SKUs across AI Native ITSM and CBS, including associated entitlement-driven flows and experiences.
• Product Hub UX Enhancements: Introduced targeted UX improvements to Product Hubs, including updates to the Resources section and guided content refinements based on leadership and customer feedback to create a more intuitive onboarding experience.
• LitJS Component Adoption in NAP: Enabled the ability to import and leverage LitJS components within NAP, helping establish a more scalable and consistent component framework.
• Admin Home Banner Updates: Updated Admin Home banners and messaging to better guide users through setup, actions, and key implementation milestones.
• Zero-Touch Implementation: Expanded Zero-Touch Implementation capabilities to deliver provisioned instances with entitled plugins and apps fully enabled, giving customers a value-ready environment from day one.
• Intelligent Configuration Update Tracking: Built on the April GA release with smarter tracking of configuration updates and changes, helping customers stay informed of new capabilities and recommended actions.
• Targeted UI Improvements: Delivered incremental UI enhancements across the experience, informed directly by customer and leadership feedback to improve usability and flow.
• Common Component Adoption Across Consoles: Continued adoption of shared/common components to create a more consistent and unified experience across consoles.
• Hierarchical Agents, Phase 1: Introduced Phase 1 of the hierarchical agent orchestration layer, laying the foundation for more intelligent, coordinated experiences.
• Impact BU Uptake: Expanded onboarding and enablement to support Impact BU adoption and uptake of the platform and implementation capabilities.

New feature:

Entitlement creation error handling - This feature uses AI to suggest resolutions when users upload the standard Excel import template for entitlement creation.

New:

• The Intent to Action agentic workflow analyzes incoming emails to automatically identify the appropriate procurement intent and create corresponding procurement case records. Procurement teams can review the generated cases and take the necessary actions to efficiently manage procurement requests.
• The Savings Opportunity Discovery agentic workflow analyzes contracts, spend data, sourcing pipeline data, and supplier performance on a recurring schedule to automatically identify a prioritized list of savings opportunities. Category managers can review each opportunity in the Now Assist Panel and create a pipeline project or dismiss the opportunity directly from the panel.

New:

• Introduced Project Q&A conversational capability to ask questions about your project in natural language and get instant, AI-generated answers directly from the project workspace, without having to manually analyze project data.
• Use Portfolio insights to analyze planning items in a portfolio plan for execution delays, delayed starts, planned versus approved date misalignments, and projects at risk. View AI-generated root causes and recommended actions to mitigate identified risks and issues.

Changed:

• Project insights now includes scope creep as an additional topic and considers work notes at the task level as additional context, providing richer and more actionable project visibility.
• Updated formatting for in-app insight widgets to improve readability and presentation within the project workspace.
• Project Risk Detection now includes clickable links in AI Rationale for quick navigation to referenced project records. Refined Mitigation plan formatting for a better reading experience.

New:

• Ability to link emails to the created cases and the supplier records
• Ability to create assessment templates and error handling while cancelling created assessments

Migrated Generate Talking Points and Text to Growth Plan skills to the Mosaic AI framework.- Added Interview Scheduling enhancements.- GenAI evaluation framework: created GenAI MATs and auto-eval runs for Job Requisition Creation and Interview Scheduling agentic flows.