AIster · The agentic age · Laws & rulings
EU rules for general-purpose AI models take effect
The first binding legal duties aimed squarely at the makers of frontier models
Key facts
- Date
- 10 Jul–2 Aug 2025
- Era
- The agentic age
- Interest
- Laws & rulings
- Threads
- Model wars, Safety & governance
- Code published
- 10 July 2025, with chapters on transparency, copyright, and safety and security
- Obligations apply
- 2 August 2025; enforcement powers from 2 August 2026
- Older models
- Models already on the market must comply by 2 August 2027
- Did not sign
- Meta; xAI signed only the safety and security chapter
On 10 July 2025 the European Commission published a voluntary Code of Practice showing how makers of general-purpose AI models, built for a wide range of tasks, can meet the EU AI Act. The Act's obligations for these providers applied from 2 August 2025, with the Commission able to enforce them, including with fines, from 2 August 2026. OpenAI, Google, Anthropic and others signed the code; Meta refused.
Sources
- European Commission, The General-Purpose AI Code of Practice, 2025-07-10
- European Commission, Guidelines for providers of general-purpose AI models, 2025-08-02
- TechCrunch, Meta refuses to sign EU's AI code of practice, 2025-07-18