Track EU AI Act compliance where your AI systems already live

The EU AI Act is now in effect for organizations that develop, deploy, or use AI systems in the EU market. Two complementary new capabilities bring governance for this regulation directly into your inventory and connect it to the AI systems and applications you already manage in your workspace. You get a structured, portfolio-wide view of where the act's considerations apply, without having to recreate data already in the inventory.

1. Workspace admins can activate the opt-in EU AI Act Management meta model extension to add fields to application fact sheets that cover risk classification, scope determination, and documentation requirements. This extension works best with the AI Governance extension, but doesn't require it to be activated.

2. The EU AI Act classification is automatically available when using the Enterprise Architecture Assistant to guide you through assessment and gap surfacing across your architecture. It answers questions grounded in the EU AI Act text with citation-backed responses, walks you through risk-tier assessments for prohibited, high-risk, limited, and minimal risk categories, including transparency obligations, and extracts structured answers from your existing documentation, asking only for what's missing and outputting a draft risk tier. It also surfaces fact sheets lacking AI Act fields, pre-fills from completed assessments, and flags suspected high-risk systems without documentation.
   All outputs are drafts for your architecture, legal, and compliance teams to review, as the assistant never asserts a compliance verdict. Responsibility for assessing and demonstrating compliance remains with your organization.