Regulatory Standards · European Union
EU Artificial Intelligence Act
EU AI Act obligations, backed by evidence of what AI does.
The AI Act sets risk-based obligations for AI providers and deployers, phased in from 2025 to 2028. Transparency rules already apply; the 2026 Digital Omnibus moved high-risk obligations to December 2027. Astragar maps requirements to controls, and Aeguard evidences what AI agents actually access and do.
Register interest in EU AI ActComing to Astragar GRC
7% / €35M
Maximum fine for prohibited AI practices, whichever is higher
Aug 2026
Transparency duties for chatbots and AI-generated content apply
Dec 2027
High-risk (Annex III) obligations apply, as deferred by the Digital Omnibus
What EU AI Act requires
The obligations, in brief.
Classify AI systems
Identify which systems are prohibited, high-risk or limited-risk.
Risk management
Risk management, data governance and human oversight for high-risk AI.
Logging & transparency
Record-keeping and transparency to users and authorities.
Accuracy & security
Robustness, accuracy and cybersecurity of high-risk systems.
How Astragar helps
From checklist to measurable risk.
01
Requirement → control mapping
Map every EU AI Act requirement to controls, systems, assets and owners.
02
Evidence, continuously
Collect and maintain evidence, with sign-off from contributor to IT auditor, CISO and executive office.
03
Exposure behind the gaps
See which vulnerabilities and sensitive data put EU AI Act controls at risk.
04
Risk in financial terms
Prioritise remediation by business and financial impact, not checklist order.
Use the deferral to build evidence now. Aeguard attributes AI-agent activity to owners and logs what agents access, with tamper-evident records.
Sign-off workflow
Evidence Contributor
→IT Auditor
→CISO
→Executive Office
Related · Government, Critical Infrastructure & AI
European Union
NIS2 →
Saudi Arabia
NCA ECC →
European Union
EU CRA →
United States
FedRAMP →
United States
CMMC →
Global
ISO 42001 →
Global
NIST AI RMF →
Find the data. Test the controls. Reduce the risk.
EU AI Act mapping is coming to Astragar GRC. Talk to us about early access and how your existing evidence already applies.
Talk to usAstragar supports compliance programmes with technical evidence and risk analysis. It does not provide legal advice or issue certifications.

