Signing the AI code doesn’t mean they like it
Google’s signing up to the EU’s new code of practice for AI, a voluntary guide that helps powerful AI models follow the upcoming AI Act.
The aim is to make sure AI is used safely, respects copyright, and is clear about how it’s trained.
Google says this will help bring high-quality, secure tools to people and businesses across Europe.
Still, the tech giant isn’t fully on board with everything.
It’s worried that parts of the Act could actually slow down progress.
Rules that might force companies to reveal trade secrets or delay approvals could end up hurting Europe’s ability to compete globally, according to Google.
In brief:
Google signs on, but flags concerns around red tape
Meta stays out, warning of overreach
EU says digital rules aren’t up for negotiation
Signing ≠ agreeing
Despite criticism from the US and some tech groups, the EU says its digital rules are final.
It’s made it clear that it won’t water them down for trade deals or outside pressure.
OpenAI and France’s Mistral have also signed the code.
Microsoft is expected to follow.
Meta, though, is not joining; it argues the code creates too much legal uncertainty and goes beyond the law itself.
Google said “we’ll sign, but don’t touch our secrets.