The European Union has published its General-Purpose AI Code of Practice on July 10, 2025, aiming to guide AI developers in complying with the EU AI Act and avoid potential penalties. The Code includes three chapters: Transparency, Copyright, and Safety and Security, with the final chapter applying only to providers of advanced models posing systemic risks, such as OpenAI’s ChatGPT, Meta’s Llama, and Google’s Gemini. It requires developers to disclose training data sources, respect copyright laws, and implement risk management frameworks. Signing the Code is voluntary, though it offers companies a pathway to demonstrate alignment with upcoming EU regulatory deadlines. The AI Act, which took effect on August 1, 2024, enforces its provisions in phases, with the next significant milestone on August 2, 2025, when requirements for general-purpose AI models come into force.
OpenAI has signed the Code, while Meta declined to do so, citing legal uncertainties and regulatory overreach. “This Code introduces a number of legal uncertainties for model developers, as well as measures which go far beyond the scope of the AI Act,” wrote Joel Kaplan, Meta’s Chief Global Affairs Officer, on LinkedIn. “We share concerns raised by these businesses that this over-reach will throttle the development and deployment of frontier AI models in Europe, and stunt European companies looking to build businesses on top of them.” Critics argue that the Code’s voluntary nature could delay legal clarity and fragment compliance efforts. The European Commission will issue further guidance before August 2 to clarify provider responsibilities.
