general-purpose AI

AI that handles many different kinds of task across domains, including many it was never specifically built for, usually because it rests on a foundation modelfoundation model, n.: A very large AI model, built once at great cost from vast amounts of text, images or code gathered mostly from the internet, that many products then reuse as their starting point, from chatbots to coding tools. pretrained on wide-ranging data and adapted through prompting or fine-tuning. Policy writing uses the phrase as a present-tense, capability-based alternative both to 'foundation model', which describes how a model is built, and to artificial general intelligenceartificial general intelligence, n.: AI that could learn and do almost any mental task a capable adult can, rather than excelling at one narrow job. Experts disagree on the exact bar and on when, or whether, it will arrive., which implies human-level breadth. The International AI Safety Report, chaired by Yoshua Bengio (first full edition January 2025), takes general-purpose AI as its object of study, covering language models, multimodal models and agents built on them.

Because it is general-purpose AI, the vendor cannot list every use in advance, so the risk assessment covers misuse too.

EU lawlegal

Under the EU AI ActEU AI Act, prop. n.: The European Union's law on artificial intelligence, the first broad AI law anywhere: it bans some uses, sets strict checks for risky ones and puts duties on the makers of the biggest AI models., a general-purpose AI modelgeneral-purpose AI model, n.: The EU's legal name for the large, adaptable AI models behind chatbots and similar tools: models that can turn their hand to many different jobs and be built into many other products. is one that shows significant generality, is 'capable of competently performing a wide range of distinct tasks' and can be built into many downstream systems or applications (Art. 3(63)); a general-purpose AI system is an AI system built on such a model and usable for many purposes, whether on its own or as a component of other AI systems (Art. 3(66)). Model providers have had transparency and copyright duties since 2 August 2025; models with high-impact capabilities, presumed above 10²⁵ floating-point operations of training compute, are models with systemic riskgeneral-purpose AI model with systemic risk, n.: The EU's label for the very largest and most capable AI models, whose failure or misuse could cause harm across society, so their makers must test them for dangers and report serious incidents. and carry further duties.