How to Prepare AI Agents for the EU AI Act
Begin with role and risk analysis, distinguish obligations applying now from amended high-risk dates, and connect readiness work to evidence and human review.
Updated
Resource Library
Practical guidance for decisions about AI agents
Understand the evidence behind an agent decision, the boundaries of an assurance record and the work required to keep both current.
For: enterprise AI owners, engineering, security, risk and governance teams
EU AI Act
Enterprise Assurance
What is AI agent certification? Process and claim limits
Understand AI agent certification, the assessment process and the questions enterprise buyers should ask about scope, issuers, evidence and reassessment.
Updated
AI agent access certification vendors and buying criteria
Compare identity, authorization, testing and formal certification offerings so your AI agent access review buys the controls and evidence it actually needs.
Updated
AI agent assurance: an enterprise evidence checklist
Learn how to review AI agent assurance evidence, assess coverage gaps and make a bounded deployment decision, with a practical enterprise checklist.
Updated
How to certify AI agent access with scoped evidence
Assess AI agent access through permission inventory, enforcement tests, approval checks and review evidence tied to an exact deployment and operating scope.
Updated
Agent Security
Prompt Injection Testing for Enterprise AI Agents
Test whether untrusted instructions can redirect an agent across its tool, data or authority boundaries, and preserve evidence that reviewers can reproduce.
Updated
AI agent credential check for identity and access
Run an AI agent credential check that verifies issuer trust, agent identity, token validity, audience, permissions and revocation before allowing access.
Updated
AI Agent Security Certification: What Evidence to Ask For
Evaluate AI agent security certification claims through scope, threat coverage, test evidence, critical findings and the conditions on a release decision.
Updated
Jailbreak testing: a controlled evaluation methodology
Design authorized jailbreak tests with clear policy criteria, representative cases, calibrated scoring, and evidence that supports a bounded release decision.
Updated
Prompt extraction: risks, testing, and practical defenses
Assess system prompt extraction by the information actually exposed. Learn how to verify leakage, remove secrets, and enforce permissions outside the model.
Updated
Prompt injection vs jailbreak: differences and risks
Understand how prompt injection and jailbreaks overlap, which boundaries they threaten, and how enterprise teams can classify findings and choose defenses.
Updated
Monitoring and Drift
Why AI Agents Need Runtime Monitoring After Assessment
Runtime monitoring connects changes in an agent and its environment to investigation, human decisions and proportionate reassessment.
Updated
Scorecards and Passports
What is an agent passport? How to read and verify one
Learn what an agent passport tells enterprise buyers, how to check its source, version, scope and status, and when to request a deeper assurance review.
Updated
Agent Trust Score: How to Read the Evidence Behind a Rating
Learn how to interpret an agent trust score, compare assessment coverage and spot critical failures, missing evidence and misleading aggregate results.
Updated
AI Agent Passport Standards: What Enterprises Should Know
Understand the standards and protocols behind agent identity, credentials and interoperability, and the evidence an enterprise agent passport still needs.
Updated
Governance
AI Agent Governance vs AI Model Governance
Model governance examines a component; agent governance must also govern actions, permissions, evidence and decisions across a changing operating system.
Updated
Put your next agent decision on a clear evidence basis
Discuss the agent, its operating boundaries and the assurance work your enterprise needs.