Google agrees to sign the EU's AI rulebook

Meta recently threw a tantrum over EU regulation, refusing to sign its AI code, slamming it as overreach, and warning Europe is headed down the wrong path.

Google is taking the exact opposite route.

Kent Walker, Google's president of global affairs, announced the company will officially sign the EU General Purpose AI Code of Practice. The move aligns Google with several other U.S. model providers, just days before August 2, when strict baseline rules for systemic-risk general-purpose AI models take effect.

Playing nice on paper, venting in public

The code of practice is technically a voluntary framework designed to help major AI labs navigate the landmark EU AI Act before its mandatory two-year compliance window closes. Heavy hitters in this tier include OpenAI, Google, Anthropic, and Meta.

Agreeing to the code comes with real friction. Developers must provide updated system documentation, avoid training on pirated content, and honor content creators' requests to opt out of training datasets.

Google isn't pretending it loves every line. While Walker acknowledged the finalized text is better than the original draft, he aired serious reservations. He warned the AI Act and its code could stall European development, pointing directly to potential departures from existing EU copyright law, slow approvals, and rules that might force companies to expose trade secrets.

Still, Europe's economic upside is simply too big to ignore. Google projects widespread AI adoption could boost the European economy by 8 percent annually by 2034, adding roughly 1.4 trillion euros.

Why it matters

For businesses and developers across Europe, Google signing means its flagship Gemini models will keep flowing into European workflows without sudden regulatory blackouts. It also avoids the fragmented rollout mess Meta users are currently navigating.

For the broader industry, Google's signature completely fractures Silicon Valley's unified front against European regulators. Meta bet on total defiance to force regulatory pushback. Google decided that retaining enterprise market share and staying in the room with the EU AI Office is worth the compliance headache.

We think Google played this smartly, even if it looks contradictory. By signing, Google secures its distribution rights and locks in institutional goodwill, while preemptively pinning any future tech stagnation squarely on bureaucratic red tape. The risk, however, is clear: agreeing to enforce opt-outs and reveal documentation creates an expensive compliance moat that could easily tighten into trade secret exposure later.