Compliance Roadmap
Interactive timeline of the EU AI Act enforcement milestones.
Interactive Regulatory Timeline
Filter by specific compliance years or responsible actors to isolate critical deadlines.
Regulatory Entry into Force
The EU AI Act formally becomes active as a binding European regulation. While initial compliance duties are not yet mandatory, this sets the baseline calendar from which all staggered transitional timelines are officially calculated.
Article 113Fundamental Rights Protection Mapping
All EU Member States must identify and establish public authorities and agencies responsible for protecting fundamental human rights under the Act's provisions. These bodies must be declared to the EU Commission.
Article 77(2)Unacceptable Risk Market Bans
Strict European market prohibitions go into active effect. AI systems deemed as presenting unacceptable societal risks (e.g., social scoring, cognitive manipulation, facial biometric scraping) are permanently banned. Organizations must also begin workplace AI literacy training.
Article 113(a) & Recital 179Codes of Practice Finalization
The European AI Office drafts the primary Codes of Practice. These documents outline the expected operational compliance metrics for providers of General-Purpose AI (GPAI) models ahead of upcoming structural audits.
Article 56(9)GPAI Model Rules & Governance Launch
Regulatory frameworks activate. Obligations for General-Purpose AI (GPAI) model providers become legally binding. Rules regarding Notified Bodies (independent conformity assessment agencies), confidentiality provisions, and penalty rules also become active.
Article 113(b)Legacy GPAI Transition Window Starts
General-Purpose AI models that were already actively deployed or placed on the European market prior to August 2025 are granted a two-year transitional compliance runway, setting their ultimate compliance deadline to August 2, 2027.
Article 111(3)Designating National Supervisory Authorities
Each EU member country must formally designate national competent authorities, identifying their respective notifying authorities and market surveillance authorities to handle local oversight and report to the central Commission.
Article 70(2)Fines & Penalty Systems Codification
Member States must draft and formalize local penalty structures and administrative fine parameters to discipline violations of the regulation, communicating the implemented systems directly to the EU Commission.
Recital 179High-Risk Post-Market Tracking Guidelines
The European Commission publishes explicit operational guidelines clarifying high-risk AI system categorizations and setting post-market tracking parameters for providers of complex systems.
Article 6(5) & Article 72(3)Annex III High-Risk Systems Application
Enforcement extends to high-risk systems under Annex III. This triggers compliance protocols for biometrics, infrastructure, employment/HR, and law enforcement applications. Additionally, all Member States must establish at least one operational national sandbox.
Article 113 & Article 57(1)Annex II Safety-Regulated AI Activation
High-risk classification standards under Article 6(1) apply. This includes AI integrated as safety components of products already regulated under EU safety directives (e.g., medical equipment, industrial machinery, transportation safety devices).
Article 113Legacy GPAI Model Alignment Deadline
The grace transition period expires. All legacy GPAI models on the market before August 2025 must now be fully compliant with transparency, documentation, and data-governance standards.
Article 111(3)AI Office Effectiveness Assessment
The Commission completes a performance and governance evaluation of the European AI Office, analyzing institutional effectiveness. Concurrently, a progress review evaluates energy-efficient model standards and voluntary industry codes of conduct.
Article 112(5) & Article 112(7)First Structural Regulatory Review
A comprehensive review of the AI Act is prepared for the European Parliament. This covers critical reassessments of Annex III high-risk boundaries, transparency measures, and the Commission's delegate powers to modify annexes.
Article 112(3) & Article 97(2)Public Sector High-Risk AI Deadline
High-risk AI systems implemented by public bodies and administrative agencies must reach complete compliance. Additionally, components of EU large-scale information platforms (Annex X) must be modernized to match the AI Act by December 31, 2030.
Article 111(2) & Article 111(1)Legislative Enforcement Performance Audit
The European Commission presents its first formal performance audit on the practical enforcement of the AI Act to legislative authorities, concluding the multi-year implementation cycle with a comprehensive evaluation report.
Article 112(13)