GPAI (general-purpose AI)
AI models built for broad, general use rather than a single task. The EU AI Act places obligations on GPAI model providers, and the Commission's enforcement powers over them were scheduled to begin August 2, 2026 (not delayed by the Omnibus).
Defined in 4 GAGE programs, which carry 6 distinct definitions of it. The wording above is taught in Business AI Transformation.
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
A category under the EU AI Act referring to AI models trained on broad data and usable across a wide range of tasks (large language models are the most prominent example). GPAI providers face specific transparency and documentation duties under the Act, in force since August 2, 2025, with a GPAI Code of Practice published July 2025 setting out how providers can demonstrate compliance.
AI models built for broad, general use rather than a single task. The EU AI Act places obligations on GPAI model providers, and the Commission's enforcement powers over them were scheduled to begin August 2, 2026 (not delayed by the Omnibus).
An AI model, under EU AI Act terminology, capable of competently performing a wide range of distinct tasks, subject to specific obligations including public training-data summary disclosure under Article 53(1)(d).
The EU AI Act's cross-cutting regulatory tier for foundation-model providers, applying regardless of the downstream application the model is used for. Obligations became enforceable 2 August 2025.
A category of AI model, defined under the EU AI Act, capable of performing a wide range of tasks; GPAI providers carry specific obligations including publishing a training data summary.
Where it is taught
The exact lessons this term appears in. The first 7 topics of every program are free with a free account.
- Why the data person just became the most consequential role in the building · Taking the Data Chair, AI Data Governance: The Data Chair
- Membership inference and extraction: what a model reveals about its training data · Poison, Leaks, and the Adversary, AI Data Governance: The Data Chair
- Transforming Across the Regulatory Map · Lab: Regional and Geopolitical AI, Business AI Transformation
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.