Google announced on July 30 its intent to join the European Union’s General Purpose AI Code of Practice, a voluntary framework based on the EU’s AI Act. The obligations take effect on August 2, 2025, with specific computational thresholds: models exceeding 10^23 FLOPs must meet standard transparency requirements, while those crossing 10^25 FLOPs are classified as systemic-risk models. Google joins OpenAI, Microsoft, Amazon, Anthropic, Mistral AI, and IBM as signatories, while Meta and several Chinese AI firms have declined to join. Europe is counting on projections suggesting responsible AI deployment could deliver an 8% economic gain for the continent, roughly €1.4 trillion annually by 2034. AI-generated text, images, and media will require clear identification, affecting everything from Google’s search results to its advertising ecosystem.
Source: Read the original article
