Story
On August 2, 2026, a major tranche of the EU AI Act (Regulation 2024/1689) becomes enforceable, bringing GPAI model provider obligations under Article 53, high-risk AI system requirements under Articles 6-51, and deployer responsibilities under Article 26 into full effect. This deadline marks the first major enforcement milestone for general-purpose AI model providers like OpenAI, Anthropic, and Google, who must maintain technical documentation, provide downstream users with compliance information, implement copyright policies, and publish training data summaries. The European Commission gains enforcement powers including the ability to levy fines up to €35 million or 6% of global annual turnover. Organizations using AI tools like ChatGPT for high-risk use cases such as recruitment screening or credit scoring face classification as high-risk AI system deployers and must establish risk management systems. rehydra reported the core details. Legiscope noted additional context. European Commission Digital Strategy added corroborating details.