EU GPAI model rules enter application: transparency and copyright duties begin for general-purpose AI providers

[Founding Archive] AI20 editorial coverage based on European Commission (Shaping Europe’s digital future) (2025-08-01). In a 1 August 2025 European Commission press release, the Commission states that Artificial Intelligence Act obligations for providers of general-purpose AI (GPAI) models have entered into application across the EU. From 2 August 2025, providers must comply with transparency and copyright obligations when placing GPAI models on the EU market; models already on the market before that date must ensure compliance by 2 August 2027. The Commission defines GPAI models for this purpose as those trained with over 10^23 FLOP and capable of generating language, and notes additional duties for systemic-risk models exceeding 10^25 FLOP, including notification and strengthened safety and security expectations. To support compliance, the Commission published guidelines on who must comply, a template for summarising training data, and confirmed — with Member States — that the GPAI Code of Practice is an adequate voluntary tool for demonstrating compliance. The release frames the package as increasing transparency, safety and accountability for AI systems on the European market. Why it matters: Regulatory clocks are now live for foundation-model providers selling into Europe. Product, legal and go-to-market teams need operational compliance programmes — documentation, copyright policy, training-data summaries and systemic-risk notifications — not slide-deck awareness. AI20 perspective: Clear public rules reduce ambiguity for responsible builders. AI20 should help members translate EU GPAI requirements into practical cross-border playbooks that preserve innovation while meeting transparency and accountability expectations. (Category: Policy; Region: European Union; Tags: EU AI Act; GPAI; regulation; transparency; copyright; systemic risk. Original source: https://digital-strategy.ec.europa.eu/en/news/eu-rules-general-purpose-ai-models-start-apply-tomorrow-bringing-more-transparency-safety-and)