UnifAI maps 55 policies to EU AI Act requirements—helping organizations translate requirements for risk management, transparency, human oversight, cybersecurity, record-keeping, and more into operational controls and documented evidence.
55 Policies Mapped to EU AI Act Requirements
UnifAI coverage includes requirements related to:
Article 15 — Accuracy, Robustness & Cybersecurity
20 Policies Mapped | Enforced
The deepest mapping in Lineaje’s EU AI Act assessment. Continuous testing and hardening controls help keep high-risk AI systems resilient against manipulation and drift.
Article 14 — Human Oversight
5 Policies Mapped | Enforced
Runtime checkpoints gate high-risk automated actions behind human review, helping keep a person accountable in the loop.
Article 12 — Record-Keeping
4 Policies Mapped | Notification
Automatic, tamper-aware logging of AI system events provides a verifiable trail for auditors and regulators.
UnifAI also maps policies to:
Article 6 — Classification Rules for High-Risk Systems
Flags whether an AI system meets high-risk classification triggers.
Article 9 — Risk Management System
Continuous risk scoring across the AI inventory supports a lifecycle risk-treatment record.
Article 13 — Transparency to Deployers
Supports structured disclosure of system behavior and limitations to downstream deployers.
Article 26 — Obligations of Deployers
Helps confirm deployer-side controls are in place before a high-risk AI system goes live.
Article 50 — Transparency Obligations
Supports disclosure controls for chatbots and generated content.
Article 53 — Obligations for GPAI Model Providers
Maps baseline documentation and disclosure controls for general-purpose AI model providers.
Article 55 — GPAI Models with Systemic Risk
Supports additional safeguards for general-purpose AI models identified as presenting systemic risk.
Article 73 — Reporting of Serious Incidents
Supports structured reporting workflows for incidents meeting the Act’s severity threshold.
Article 86 — Right to Explanation
Supports controls that surface the reasoning behind an automated decision when an individual requests it.
Annex III — High-Risk AI Systems
Classification controls help identify use cases subject to higher-risk obligations.
Annex IV — Technical Documentation
Supports the standardized technical documentation required for high-risk AI systems.
UnifAI connects regulatory obligations to the controls organizations use to govern AI:
Regulatory Requirement → Policy Mapping → AI Control → Enforcement or Notification → Documented Evidence
This gives organizations visibility into not only which EU AI Act requirements are addressed, but how those requirements connect to policies, controls, enforcement, and evidence across operating AI environments.
DISCOVER — Know What AI You Need to Govern
Identify AI assets across teams and environments to establish visibility into the systems that may introduce EU AI Act requirements.
DERIVE — Map Requirements to Policies and Controls
Assess applicable EU AI Act requirements article by article and map covered obligations to policies with documented,
verified evidence.
DEFEND — Apply Guardrails Across AI Systems
Operationalize governance through enforceable and notification-based policies supporting areas such as human oversight, cybersecurity, transparency, monitoring, and data security.
Understanding what the EU AI Act requires is only the beginning. Organizations need to know which AI systems are affected, what obligations apply, whether the appropriate controls are in place, and whether they can prove those
controls are working.
UnifAI connects those requirements to the AI systems organizations govern—turning regulatory obligations into mapped policies, operational controls, enforcement, and documented evidence.
See how UnifAI helps you discover the AI you need to govern, map EU AI Act requirements to policies and controls, apply enforceable guardrails, and continuously maintain the evidence behind your compliance program.
*Coverage reflects Lineaje’s internal AI Policy Compliance Coverage assessment. “Covered” means at least one policy has documented, verified evidence against the applicable clause, article, or section. Coverage is limited to controls assessed by Lineaje and does not constitute a general assertion of regulatory compliance.