Meta declines to sign EU’s AI agreement, calling it ‘overreach’

Meta has announced that it will not sign the European Union’s (EU) code of practice for its artificial intelligence (AI) Act, just weeks before the bloc’s rules for general-purpose artificial intelligence (GPAI) models are set to take effect.

The Mark Zuckerberg-led company raised concerns about the framework’s legal implications and its broader impact on innovation.

Meta’s chief global affairs officer, Joel Kaplan, shared the company’s stance in a post on LinkedIn, stating, “Europe is heading down the wrong path on AI. We have carefully reviewed the European Commission’s Code of Practice for general-purpose AI (GPAI) models and Meta won’t be signing it.”

“This Code introduces a number of legal uncertainties for model developers, as well as measures which go far beyond the scope of the AI Act,” he noted.

The code of practice, released earlier this month, is intended to guide companies in complying with the EU’s landmark AI legislation.

It calls for transparency in AI development, including detailed documentation, the prohibition of using pirated content for training data, and mechanisms for content owners to opt out of having their work used.

The AI Act, which takes a risk-based approach, outright bans certain AI applications deemed to pose “unacceptable risks,” such as social scoring and manipulative behaviour.

It also imposes strict rules on “high-risk” systems, particularly those involving biometric data and sectors like hiring and education.

Despite pushback from major tech companies including Meta, Microsoft, Google, and Mistral AI, the European Commission is standing firm on its regulatory timeline.

New requirements for AI models with systemic risk will take effect on August 2, 2025, with a three-year window for full compliance.