Second Draft of General-Purpose AI Code of Practice Published

The European Union has unveiled the second draft of the General-Purpose AI Code of Practice, a critical framework aimed at supporting providers of general-purpose AI models in achieving compliance with the forthcoming AI Act. This milestone document builds upon the initial draft released in November 2024, integrating feedback from over 1,000 stakeholders, including Member States, international experts, and industry representatives.

The updated draft provides comprehensive best practices to guide AI providers through every stage of the lifecycle of their models, from development to deployment and monitoring. By addressing compliance challenges proactively, the Code seeks to ensure that AI systems align with the stringent regulatory expectations of the AI Act, which takes effect on August 2, 2025, and applies to models released after this date.

One of the standout features of the second draft is its forward-thinking approach to compliance. The document emphasizes the importance of embedding ethical considerations, transparency, and accountability into AI models from the outset. These principles are essential for fostering public trust in AI technologies while mitigating potential risks associated with their widespread adoption.

Stakeholders are encouraged to provide feedback on the second draft by January 15, 2025. The EU plans to incorporate these contributions into a third and final draft, scheduled for release on February 17, 2025. This collaborative effort underscores the EU’s commitment to creating a balanced regulatory framework that supports innovation while safeguarding societal values.