EU AI Act Compliance
for AI Agents
Automatically map your agent's behavior to EU AI Act requirements. Generate audit-ready compliance reports before the August 2026 deadline.
What the EU AI Act Requires
High-risk AI systems must meet specific requirements for human oversight and robustness. Inkog checks these automatically.
Article 14: Human Oversight
Mandatory for high-risk AI systems- Human approval gates before high-impact actions
- Ability to override or halt AI decisions
- Clear escalation paths for edge cases
- Audit trails of AI-human interactions
Inkog checks: Missing human approval gates, autonomous high-risk actions, missing override mechanisms
Article 15: Robustness
Accuracy and reliability requirements- Resilience against errors and faults
- Protection against adversarial inputs
- Bounded resource consumption
- Graceful degradation under load
Inkog checks: Infinite loops, prompt injection paths, unbounded API calls, missing error handling
From Code to Compliance Report
One command generates a compliance report mapping your agent's behavior to EU AI Act requirements.
Run the scan
Point Inkog at your agent code with the eu-ai-act policy flag. No configuration needed.
Get your report
Receive a structured compliance report mapping findings to specific Articles and requirements.
Fix gaps
Address identified gaps with actionable suggestions. Re-scan to verify compliance.
Frequently Asked Questions
When does the EU AI Act take effect?
The EU AI Act was published in August 2024. Most provisions, including requirements for high-risk AI systems (Articles 14, 15), become enforceable in August 2026. Organizations deploying AI agents in the EU need to be compliant by this date.
Does the EU AI Act apply to AI agents?
Yes. AI agents that make autonomous decisions, interact with users, or perform actions in regulated domains are likely classified as high-risk AI systems under the Act. This includes customer-facing chatbots, automated decision systems, and multi-agent workflows.
What does Inkog check for EU AI Act compliance?
Inkog maps your agent's behavior to Article 14 (Human Oversight) and Article 15 (Robustness) requirements. It checks for missing human approval gates, unbounded autonomous actions, prompt injection vulnerabilities, infinite loops, and missing error handling.
Can Inkog generate compliance reports?
Yes. Running `inkog scan . --policy eu-ai-act` generates a structured compliance report that maps each finding to specific EU AI Act articles and requirements. This report can be used as part of your compliance documentation.
Start your EU AI Act compliance audit
Free for developers. Results in 60 seconds.