• Law4Startups
  • Posts
  • ⚖️ Google to sign EU's AI Code of Practice

⚖️ Google to sign EU's AI Code of Practice

Google Agrees to Voluntary AI Rules as EU Law Looms

Google has confirmed it will sign the European Union’s new voluntary Code of Practice for General-Purpose AI, aligning itself with the bloc’s AI Act just days before key rules for high-risk AI systems go into effect on August 2. The Code of Practice is intended to help companies prepare for the legally binding obligations under the AI Act, particularly those developing large models deemed to pose “systemic risk.” While Meta refused to sign the code, calling it regulatory overreach, Google noted in a blog post that although it has concerns, the final draft of the code represents progress over earlier proposals.

EU Regulation vs. Industry Hesitations

Google’s decision to endorse the code indicates a more pragmatic approach than Meta’s outright rejection. It demonstrates a willingness to operate within the EU’s risk-based regulatory framework, despite ongoing concerns about compliance burdens and the potential exposure of trade secrets. The AI Act, the first of its kind worldwide, bans certain harmful AI uses outright and enforces strict documentation, copyright, and transparency requirements on high-risk applications. By backing the code, Google likely aims to influence ongoing regulatory discussions from within rather than risk exclusion from one of the globe’s largest markets. This divide among tech giants also reflects internal strategic differences — some see regulation as a competitive advantage; others view it as a barrier to innovation.

Preparing for a Compliance-First Future

Startup founders building in the AI space, especially in Europe or with European users, should closely monitor how the EU’s AI Act is enforced over the next two years. Though the current code is voluntary, it indicates the future direction of legal requirements. Founders should begin by mapping out their model’s risk classification and developing documentation, auditability, and dataset sourcing protocols accordingly. If your startup fine-tunes, licences, or builds on large language models, understand how your upstream providers are managing EU compliance — because liability may flow downwards. Finally, expect more enterprise clients to inquire about AI regulatory compliance during procurement and due diligence. Early adoption of these best practices could provide a competitive advantage.

In addition to our newsletter we offer 60+ free legal templates for companies in the UK, Canada and the US. These include employment contracts, investment agreements and more