Hot on the heels of the General-Purpose AI Code of Practice, the European Commission has now published its guidance for general purpose AI models.
The guidance clarifies the scope of the obligations for providers of general-purpose AI models under the AI Act. The timeline for the obligations is set out below.
The guidance defines general-purpose AI models as those trained with computational resources exceeding 1023 floating point operations and capable of generating language (text or audio), text-to-image or text-to-video. It also outlines what constitutes a “provider” and ”placing on the market”, and clarifies exemptions for models released under a free and open-source licence that meet transparency conditions. The guidance also explains the implications of adhering to the General-Purpose AI Code of Practice and outline Commission’s expectations for compliance.
Finally, it also clarifies specific obligations for providers of the most advanced or impactful general-purpose AI models, particularly those posing systemic risks, such as risks to fundamental rights, safety, and potential loss of control over model, who must assess and mitigate these risks.
Its key points are:
- Clear definitions: The guidelines introduce straightforward technical criteria, for instance for when an AI model is considered “general-purpose.” This helps developers understand if they need to comply with the obligations under the AI Act.
- Pragmatic approach: The guidelines are designed to be pragmatic. For instance, they clarify that only those making significant modifications to AI models need to comply with the obligations for providers of general-purpose AI models, not those making minor changes.
- Exemptions for open-source: The guidelines also clarify under what conditions providers of open-source AI models are exempt from certain obligations, to promote transparency and innovation.
Timeline for application
From 2 August 2025, the obligations for providers of general-purpose AI models enter into application. Providers of general-purpose AI models placed on the market after this date must comply and are expected to informally collaborate with the AI Office’s technical staff. In particular, providers of the most advanced models, that is, those that pose systemic risks, are legally obliged to notify the AI Office of these models. The AI Office will support providers in their compliance, especially Code of Practice signatories.
From 2 August 2026, the Commission’s enforcement powers enter into application. The Commission will enforce compliance with the obligations for providers of general-purpose AI models, including with fines.
By 2 August 2027, providers of general-purpose AI models placed on the market before 2 August 2025 must comply.
