Agentic AI and generative AI (Gen AI) features for Integrated Risk Management (IRM), powered by Now Assist, enhance analysts' productivity, team coordination, and overall efficiency.
Issue summarization
The Now Assist for IRM application can quickly analyze an issue record, including the description, activity log, and remediation tasks, and then generate a concise summary to help you quickly understand.
Issue resolution agentic workflow
Optimize GRC issue resolution is an agentic workflow in the Integrated Risk Management application that enables issue managers and analysts to resolve GRC issues through the Now Assist panel. This workflow leverages AI agents to suggest targeted solutions for key steps during the resolution process.
Risk assessment summarization
Risk assessment summary generation uses generative AI to assist assessors during submissions and provide approvers and stakeholders with a quick overview of assessment progress and key insights.
Control objective rationalization and deduplication
The control objective rationalization and deduplication process leverages generative AI to identify similar or duplicate control objectives, enabling teams to consolidate them by retaining one, retiring duplicates, or merging them into a common control objective, enhancing overall efficiency across compliance operations.
Risk event summarization
The risk event summarization feature helps risk managers quickly understand the context and key details of a risk event, even when multiple coordinators have been involved over time. It streamlines onboarding for new assignees and improves decision-making efficiency for approvers, especially in multi-level approval workflows, by providing concise AI-generated summaries of each event.
Rationalization of control objectives
Skill for creating a common control objective as part of the rationalization process for control objectives.
Control recommendation on regulatory alert
Skill for recommending controls for an incoming regulatory alert.
Issue submission agent
The issue submission AI agent is now available in Employee Center, enabling employee users to report issues through a guided conversational experience. The agent helps structure the issue and recommends relevant controls, entities, and policies based on the input provided. This ensures that the issue is well-defined and enriched with contextual information before submission.
Suggest potential risks agentic workflow
AI-driven risk identification analyzes entity context and surfaces relevant risks from internal libraries, industry patterns, and optional external sources. It generates a single, ready-to-triage risk list, reducing manual effort and duplication while ensuring comprehensive coverage.
Control objective impact analyzer
The citation-driven impacted control objective identifier uses generative AI to analyze changes in citations and identify control objectives that should be reviewed based on the modified citation content. It intelligently matches changed citation requirements with existing control objectives to suggest potential updates, helping teams maintain alignment between citations and control objectives.
Control objective change agent
This capability introduces an AI-driven process to automatically update a control objective’s description and supplemental guidance whenever its associated citations are modified. Leveraging agentic AI, it analyzes citation changes, identifies their impact on the control objective, and generates precise, contextually relevant updates.
New
- Added contextual follow-up questions in the Risk Identification Agent to improve the accuracy of suggested risks.
- Introduced a review and feedback step that allows users to update or remove risks before they are created.
- Enhanced security for the Risk Assessment Summarization skill through role masking.
Fixed
- Fixed an issue that could result in null reference errors in risk assessment summaries.
- Improved the overall performance of the Risk Identification Agent.
- Fixed an issue where the Issue Resolution Agent displayed remediation tasks twice, with formatting issues across multiple LLMs.
- Fixed an issue where the Issue Submission Agent did not create the issue triage record with the correct details.
- Fixed oversized remediation task confirmation header formatting for AWS Claude.
- Fixed role mask changes and security issues.
Changed
- UCF control objectives are now part of the rationalization process.
Requirements:
- Now Assist for Platform must be installed.
- GRC: Risk Management Workspace or GRC: Compliance Management Workspace (minimum version 21.x).
- AI Search must be enabled on the instance.
- Generative AI skills and agentic workflows must be activated through Now Assist Admin or Agentic AI Studio.