EU AI Act: Provisional Deal Grants Businesses Two-Year Extension on High-Risk Compliance
Breaking: EU Lawmakers Agree to Push Back AI Act Deadlines
European Union member states and the European Parliament struck a provisional deal early Thursday to extend the compliance deadlines for high-risk AI systems, giving enterprises two more years to adapt. The revised timeline pushes the deadline for stand-alone high-risk systems to December 2, 2027, and for AI embedded in products covered by sectoral safety rules to August 2, 2028. The original deadline was August 2, 2026.

The agreement must still be formally adopted by both the Parliament and the Council before it becomes law. The co-legislators aim to finalize this step before August 2; until then, the original deadline remains in effect.
Quote from EU Official
“Today’s agreement on the AI Act significantly supports our companies by reducing recurring administrative costs,” said Marilena Raouna, Cyprus’s deputy minister for European affairs, speaking on behalf of the Council. Cyprus holds the rotating presidency of the Council, which represents the EU’s 27 member states.
Background: Why the Deal Matters
The breakthrough comes just nine days after previous negotiations collapsed without agreement. The AI Act, first proposed in 2021, aims to regulate artificial intelligence based on risk levels. High-risk systems—such as those used in critical infrastructure, employment, or law enforcement—face the strictest requirements, including conformity assessments and transparency obligations.
Industry groups had warned that the original 2026 deadline was unrealistic, citing the complexity of compliance. The provisional deal addresses these concerns by removing overlapping rules for AI used in machinery products, ensuring they follow only sectoral safety laws while maintaining equivalent health and safety protections.
Key Details of the Provisional Agreement
The deal also narrows the definition of a “safety component” under the AI Act. AI features that merely assist users or improve performance will not automatically be classified as high-risk, provided their failure does not create health or safety risks.
For broader sectors—such as medical devices, toys, lifts, machinery, and watercraft—the co-legislators agreed on a mechanism to resolve overlaps between the AI Act and existing sectoral laws. This is intended to reduce regulatory fragmentation and administrative burden.
Timeline Adjustments
- Deadline for member states to set up AI regulatory sandboxes: pushed back by one year to August 2, 2027.
- Watermarking obligations on AI-generated content: will apply earlier than the Commission proposed—December 2, 2026 instead of February 2, 2027.
Relief for Mid-Size Firms
Mid-size companies gain more breathing room. Exemptions previously reserved for small and medium-sized enterprises (SMEs) now extend to small mid-cap companies, defined as firms with fewer than 500 employees. The deal also clarifies that the EU’s AI Office will supervise general-purpose AI systems centrally, while national authorities retain responsibility in areas like law enforcement, border management, judicial authorities, and financial institutions.

Another Key Quote
“With this agreement, we show that politics can move just as quickly as technology,” said Arba Kokalari, the Parliament’s co-rapporteur for the Internal Market and Consumer Protection Committee.
What This Means for Businesses and Compliance
The extended deadlines give companies two more years to implement high-risk AI compliance measures, reducing immediate pressure and allowing for more thorough preparation. For mid-size firms, the expanded exemptions could lower costs and simplify reporting requirements.
However, the earlier watermarking timeline means AI content generators—such as large language models—must label outputs sooner than expected. Companies should prepare for the dual challenge of earlier watermarking but later high-risk compliance. The centralization of general-purpose AI oversight under the EU AI Office also signals a shift toward more streamlined enforcement.
Next Steps
The provisional deal requires formal adoption by both the European Parliament and the Council. The co-legislators intend to complete this process before August 2, 2026. Until then, the original deadline applies as drafted. Stakeholders are advised to monitor the legislative timeline closely and begin aligning with the revised dates.
Related Articles
- Avoiding Algorithmic Overreach: A Tutorial on Proper Grant Evaluation from the DOGE Ruling
- LinkedIn Faces EU Legal Challenge Over Paywalled Profile Visitor Data – NOYB Claims GDPR Violation
- Behind the Scenes: Elon Musk's Last-Minute Settlement Attempt and Fiery Text Before OpenAI Trial
- Purdue Pharma Shutdown Approved: Judge Sentences OxyContin Maker to Dissolution
- Musk vs. Altman: The Legal Showdown Over OpenAI's Mission
- Exploring the OpenAI Smartphone Buzz: Our Take on 9to5Mac Daily's Top Stories
- EU Agrees to Loosen AI Act Timelines: High-Risk Compliance Deadlines Extended
- 9to5Mac Daily April 30, 2026: OpenAI Smartphone Buzz and Top Apple News