Overview
Meta has declined to sign the European Union's voluntary Code of Practice for general-purpose AI models, citing legal uncertainties and concerns about regulatory overreach. This decision sets Meta apart from other companies as the EU prepares to enforce its landmark AI Act.
Meta’s Position on EU AI Code of Practice
- Meta will not sign the EU’s voluntary AI Code of Practice, according to global affairs chief Joel Kaplan.
- The company cites legal uncertainties and believes the Code introduces requirements beyond the scope of the AI Act.
- Meta warns that the EU’s approach could limit the development and deployment of advanced AI models in Europe.
Details of the EU AI Code of Practice and AI Act
- The EU’s Code of Practice, introduced July 10, 2025, aims to guide companies in complying with upcoming AI Act rules.
- Signing the Code is voluntary, but signatories will have reduced administrative burdens and increased legal certainty.
- The AI Act requires general-purpose AI providers to disclose information about training and security risks and comply with copyright laws.
- Non-compliance with the AI Act can result in fines of up to seven percent of annual sales.
Industry Response and Regulatory Landscape
- OpenAI has publicly stated its intention to sign the EU’s Code of Practice.
- Over 45 companies, including Airbus, Mercedes-Benz, Philips, and ASML, have asked the EU to delay the AI Act’s implementation due to compliance concerns.
- The EU’s regulatory stance contrasts with the US, where the government is currently removing AI regulatory roadblocks.
- Meta’s refusal follows previous fines under EU regulations and aligns with US positions favoring less restrictive tech policies.
Key Dates / Deadlines
- August 2, 2025: AI Act rules come into force in the EU for general-purpose AI providers.