EU AI Act
EU AI Act high-risk mapping.
TrustScope is strongest on runtime risk management, record-keeping, transparency, and oversight controls required for high-risk systems.
Last verified mapping date: February 19, 2026
Enforcement milestones continue through 2026-2027. Penalties can reach up to 7% of global annual revenue for prohibited-practice violations.
| Article | Requirement | Status | TrustScope Evidence |
|---|---|---|---|
| Article 9 | Risk management system | Ready | Detection coverage, severity scoring, and policy response logs |
| Article 10 | Data governance and quality | Partial | Runtime data quality and leakage checks; training-data governance is out of scope |
| Article 11 | Technical documentation | Ready | Agent inventory, tool maps, and control records |
| Article 12 | Automatic record-keeping | Ready | Hash-chained logs and per-trace receipts |
| Article 13 | Transparency and instructions for use | Ready | Evidence exports and operator-facing governance records |
| Article 14 | Human oversight and intervention | Ready | Approval gates, kill-switch controls, and intervention events |
| Article 17 | Quality management system | Partial | Strong runtime evidence; organizational QMS processes remain customer-owned |
Evidence, not legal determination
TrustScope provides governance evidence, not legal conclusions. EU AI Act readiness requires legal and compliance review.