AI Regulation: Meta’s Bold Refusal Challenges EU’s Landmark AI Act

Key Points

1. Meta's chief global affairs officer, Joel Kaplan, criticized the EU's approach, stating that the Code of Practice introduces requirements beyond the AI Act's scope.

2. The EU AI Act is a risk-based framework for AI regulation, but Meta opposes the voluntary code, which includes provisions like mandatory documentation updates and restrictions on training data.

3. The contested code applies to general-purpose AI models with systemic risk, such as those from OpenAI, Anthropic, Google, and Meta.

4. Other major tech companies, including Alphabet, Microsoft, and Mistral AI, have also lobbied the European Commission to delay the rules.

5. The EU remains firm on its timeline, with compliance required by August 2, 2027, for companies with affected models on the market before August 2 of this year.