Regulation
GPAI (General-Purpose AI)
EU AI Act category for general-purpose AI models (e.g. GPT, Claude, Gemini) with broad capabilities; documentation and transparency duties apply from August 2025.
GPAI (General-Purpose AI) is the EU AI Act regulatory category for general-purpose AI models. Article 3(63) defines such a model as one that “displays significant generality and is capable of competently performing a wide range of distinct tasks.” In practice this captures large foundation models such as GPT, Claude, Gemini, and Llama.
The Act sets two tiers. Every GPAI provider must keep technical documentation, publish a summary of training data, respect copyright, and pass information to downstream integrators. GPAI with systemic risk — a model whose cumulative training compute exceeds 10²⁵ FLOPs (Article 51) — additionally faces model evaluations, risk assessment, red-teaming, and serious-incident reporting.
These obligations began applying on 2 August 2025, with a voluntary Code of Practice (published 10 July 2025) offered as a compliance route. The Commission’s power to issue fines takes effect in August 2026, and models placed on the market before August 2025 must comply by August 2027.