European AI Firms Reject Calls to Slow Model Development

European AI companies and officials are rejecting Anthropic CEO Dario Amodei's call to slow advances in model capabilities, arguing that the proposal could protect dominant US developers, Reuters reports. Mistral AI said some established companies were using safety concerns to support rules that favor them over competitors.

Amodei proposed that leading developers coordinate on safety standards, accept permanent independent evaluators and receive an exemption from antitrust rules for this work. He said voluntary coordination was needed because AI capabilities were advancing faster than governments could regulate them.

French Finance Minister Roland Lescure said slowing other developers would serve the interests of companies already in front. Germany's digital affairs ministry called a halt unrealistic, while the country's AI industry association said European companies should focus on closing the technology gap. Hugging Face CEO Clement Delangue also opposed slowing development but supported independent evaluators inside AI companies.