Google Signs EU's General Purpose AI Code of Practice Amid Rising Compliance Burdens
Google has joined the European Union's General Purpose AI Code of Practice, a voluntary framework behind the bloc's sweeping AI Act. The code covers documenting training data, ensuring copyright adherence, and mitigating risks for advanced AI systems.
The GPAI Code of Practice was finalized on July 10, 2025, and its obligations kick in on August 2, 2025. Models exceeding 10 to the 23rd FLOPs must meet standard transparency requirements, while models crossing the 10 to the 25th FLOPs threshold are classified as systemic-risk models with even heavier compliance burdens.
Google's Gemini models fall squarely within the code's scope, and the tech giant pointed to projections suggesting responsible AI deployment could deliver an 8% economic gain for Europe, roughly €1.4 trillion annually by 2034.
The training data documentation requirements could reshape the economics of decentralized AI, giving projects that can demonstrate provable, transparent data sourcing a regulatory advantage over competitors that can't.