--- sourceDocument: Australia Financial Services Operations sourceDocumentLink: https://servicenow-prod.fluidtopics.net/r/de-DE/financial-services-operations Release : - australia ft:locale : - de-DE ft:publication_title : - Australia Financial Services Operations ft:clusterId : - finso bundleId : - finso workflow : - Customer and Industry --- # Help resolve friendly fraud disputes agentic workflow # Help resolve friendly fraud disputes agentic workflow {#ariaid-title1} * Freigeben Version: Australia * * Aktualisiert 12. März 2026 * * ![](https://www.servicenow.com/docs/portal-asset/ico-clock) 3 Minuten Lesedauer Use this agentic workflow to assist human agents with analyzing friendly fraud cases, selecting a course of action, and drafting a decision response to customers. ## Help resolve friendly fraud disputes overview {#friendly-fraud-agentic-ai-workflow__section_m4v_qds_t2c} The Help resolve friendly fraud disputes agentic workflow detects friendly fraud through automated checks, guides human agents toward appropriate actions, and helps draft responses for customers. For information on modifying the agentic workflow, see [Configuring agentic workflows in Financial Services Operations](https://servicenow-prod.fluidtopics.net/tH7kse51bS8tjcZkuRCgag "Activate and modify agentic workflows for Now Assist for FSO in AI Agent Studio."). ## Help resolve friendly fraud disputes agentic workflow {#friendly-fraud-agentic-ai-workflow__section_dt2_fhg_52c} Leverage the outcomes from the friendly fraud detection to produce a recommended course of action for the human agent. To access the agentic workflow: 1. Navigate to AllAI Agent StudioCreate and manage. 2. Select Help resolve friendly fraud disputes. {#friendly-fraud-agentic-ai-workflow__ol_z52_pp2_m2c} The Help resolve friendly fraud disputes agentic workflow is inactive by default. To activate it: 1. In the Help resolve friendly fraud disputes agentic workflow, select Define trigger. 2. Select the agentic workflow trigger in the list. 3. On the Edit trigger form, set Active to true. {#friendly-fraud-agentic-ai-workflow__ol_uzb_fpk_1fc} ## Triggers for Help resolve friendly fraud disputes agentic workflow {#friendly-fraud-agentic-ai-workflow__section_rvm_vhg_52c} In the Define trigger screen, the agentic workflow begins executing when the trigger meets the following conditions: * Set the Target table field to sn_bom_credit_card_disputes_task. * Set the Run as field value to Assigned to. * Set the appropriate conditions. {#friendly-fraud-agentic-ai-workflow__ul_cbh_qhr_m2c} Wichtig: In the Edit trigger form, confirm that the Active option is turned on to trigger the AI agent autonomously. ## Friendly fraud AI agent {#friendly-fraud-agentic-ai-workflow__section_svq_3jg_52c} The following tables list the agent that is used in the Help resolve friendly fraud disputes agentic workflow. Wichtig: In the Define availability screen for the AI agent, make sure that the Status field is enabled to activate the AI agent. {#friendly-fraud-agentic-ai-workflow__table_zjv_wjg_52c__entry__2} | AI agent | AI agent role | |-|-| | Friendly fraud AI agent | Assists dispute agents in resolving friendly fraud disputes by providing transaction details, recommended actions, and compliance guidelines. It also helps draft a detailed response for the customer, explaining the reasons for any rejection in further dispute processing. | [Tabelle : 1. AI agents and their roles in the Help resolve friendly fraud disputes agentic workflow] {#friendly-fraud-agentic-ai-workflow__table_zjv_wjg_52c} ## Tools mapped to Friendly fraud AI agent {#friendly-fraud-agentic-ai-workflow__section_vt4_vkg_52c} {#friendly-fraud-agentic-ai-workflow__table_q5x_ykg_52c__entry__4} | Tool type | Execution mode | Name | Description | |-|-|-|-| | Flow actions | Autonomous | Check if friendly fraud task valid for further processing | Checks if the friendly fraud task is valid for further processing. | | Flow actions | Autonomous | Update friendly fraud resolution action taken | Updates the friendly fraud resolution action taken, action taken reason, and decline communication template that was generated for the user. While passing the value for the Action taken input, it passes the values only in lowercase letters; otherwise, it results in errors. | | Flow actions | Autonomous | Get recently closed friendly fraud tasks with overrides | Retrieves the recently closed friendly fraud tasks with overrides where the recommended action and the action taken don't match. The tool gives historical records that provide the final action and reason where the recommended action is overridden. | | Flow actions | Autonomous | Update friendly fraud recommended action and recommended action reason | Updates the friendly fraud's recommended action and recommended action reason based on the knowledge article. While passing the values for the input "Recommended action", it passes the values only in lowercase letters; otherwise, it results in errors. | | Search retrievals | Autonomous | Get knowledge article | Provides a defined set of rules to resolve friendly fraud with the help of an AI agent. It provides access to specific, standard documentation on preset organizational or IT procedures. For example, a knowledge base article. This tool isn't designed for reviewing knowledge article roles or restrictions. This tool is to be run only one time to find any relevant knowledge articles. | | Subflows | | Get friendly fraud task details | Gets friendly fraud task details. | [Tabelle : 2. Tools used in the Friendly fraud AI agent] {#friendly-fraud-agentic-ai-workflow__table_q5x_ykg_52c} ## Resolving the friendly fraud dispute {#friendly-fraud-agentic-ai-workflow__section_exb_bmg_52c} For more information on using the Help resolve friendly fraud disputes agentic workflow, see [Resolve friendly fraud by using agentic AI](https://servicenow-prod.fluidtopics.net/g6oR~TSvwxFt3nxiIETxEw "Resolve friendly fraud by using the friendly fraud AI agent to determine the actions to take for the disputed amount, the customer relationship, and assessing the outcome of the detection logic to evaluate the results.").