EU AI Act Deadlines & Timeline
The EU AI Act has been in force since August 2024 – but obligations apply in phases depending on system type. Art. 50 transparency obligations apply from 2 August 2026. The main deadline for high-risk AI systems under Annex III was postponed to 2 December 2027 by the Digital Omnibus; product-integrated AI under Annex I has until 2 August 2028.
Key deadlines at a glance
Regulation enters into force
The EU AI Act officially entered into force. No immediate obligations for companies yet, but the starting gun for the transition phases.
Bans for unacceptable risk
Title II of the EU AI Act applies: prohibited are AI systems for social scoring, subliminal manipulation, biometric categorisation by sensitive characteristics and real-time remote identification in public.
GPAI models (general-purpose AI)
Obligations for providers of General-Purpose AI (GPAI) models take effect: transparency obligations, technical documentation and – for systemic risk – enhanced risk mitigation measures.
Transparency obligations & general applicability
Art. 50 transparency obligations apply (e.g. labelling of AI interactions and AI-generated content); the governance structure and penalty framework become applicable. The main deadline for high-risk AI under Annex III was postponed to December 2027 by the Digital Omnibus (see below).
High-risk AI under Annex III (Digital Omnibus)
Latest application date for high-risk AI systems under Annex III: technical documentation, conformity assessment, EU database registration, human oversight and ongoing monitoring. The Digital Omnibus moved this main deadline from August 2026 to 2 December 2027.
Product-integrated AI (Annex I)
AI systems embedded as safety components in products under Annex I of the EU AI Act (e.g. medical devices, machinery) have until 2 August 2028. This applies to already-marketed products that require a new conformity assessment.
What must companies complete by December 2027?
- Inventory all AI systems and determine risk classes
- Identify high-risk AI systems under Annex III
- Create technical documentation according to Annex IV
- Conduct or commission conformity assessment procedures
- Establish internal governance and human oversight mechanisms
- Register high-risk AI in the EU database
How long does preparation take?
Experience shows: companies with multiple high-risk AI systems typically need 3–6 months for full compliance preparation. Anyone who has not yet started should begin immediately.
Frequently asked questions about EU AI Act deadlines
What happens if my company misses the deadline?
From each respective date (Art. 50 transparency: August 2026; high-risk under Annex III: December 2027), market surveillance authorities can take action. Sanctions range from market withdrawals to fines of up to 3% of global annual turnover (for deployer violations) or €15 million or 3% for more serious violations.
Which deadline applies to which AI systems?
Art. 50 transparency obligations apply from August 2026. The main deadline for high-risk AI under Annex III is – postponed by the Digital Omnibus – 2 December 2027. Product-integrated AI under Annex I (e.g. medical devices, machinery) has until August 2028. AI systems with minimal or limited risk have fewer requirements and no hard deadline in this sense.
What about existing AI systems that were in operation before 2024?
For already deployed high-risk AI systems, an extended transition period applies. However, systems that are substantially modified before 2 December 2027 must fully comply with the new requirements.
Do SMEs have to meet the same deadlines?
Essentially yes, but the EU AI Act includes simplified procedures and reduced requirements for SMEs. The same deadlines still apply – just the effort for some obligations is lower.
Related topics
Start compliance preparation now
SimpleAct guides you through all steps of EU AI Act compliance: inventory, risk classification, documentation and audit export – structured and on schedule.
Get started for free