Under the EU AI Act, the European Commission – and now the AI Office – may request that any developer-provider of general-purpose AI models provide the documentation drawn up in accordance with Article 53 (obligations of providers) and Article 55 (obligations for systems with systemic risk).

The AI Office – the EU’s regulatory body tasked with monitoring compliance with the AI Act – is primarily responsible for overseeing and enforcing rules related to these general-purpose AI models, particularly regarding their safety and trustworthiness. In this role, it may also assess any additional information necessary to prove compliance with the EU AI Act.