Mistral, Proton reject Anthropic call to slow AI development

September 19, 2026 at 12:43 PM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

European AI developers Mistral and Proton rejected Anthropic CEO Dario Amodei's call to slow advanced model development. They accused US rivals of exploiting safety fears to entrench market dominance. The dispute comes as EU officials push for digital sovereignty and reduced dependence on American AI.

Key Facts

Mistral accused US market leaders of using safety concerns to push regulation that benefits themselves.
Anthropic CEO Dario Amodei called for slower AI development and relaxed antitrust rules to allow joint safety measures.
Proton said Anthropic's initiative serves its own interests and aims to keep users dependent on its services.
Hugging Face opposed slowing development but supported adding independent evaluators to AI labs.
The European Commission said existing EU AI rules let Europe play a key role in minimizing risks.

European Pushback

Mistral, a leading European AI developer, said some market leaders are using the moment to strengthen their positions by calling for regulation that gives them an advantage over competitors. The French company added that the threats cited by Anthropic have been known for several months. Proton stated that Anthropic's initiative is driven by its own interests and a desire to keep users dependent on its services. Hugging Face, a French-American company, said development should be accelerated rather than slowed, though it supported Amodei's idea of adding independent evaluators to AI labs.

Anthropic's Proposal

Anthropic CEO Dario Amodei called on developers of advanced models to reduce the pace of AI improvement and argued for relaxed antitrust rules so leading labs could jointly develop safety measures. The proposal gained support from SpaceX head Elon Musk and OpenAI CEO Sam Altman. Amodei said leading developers should voluntarily agree on safety standards and accept oversight by permanent independent experts. European officials took a more skeptical stance, with German representatives saying Europe needs digital sovereignty and to overcome its technological lag.

EU Regulatory Stance

The European Commission said existing AI-related rules in the region allow Europe to play a key role in minimizing risks. European officials argued that a scheme where a few large companies define what is dangerous, choose inspectors, and demand antitrust exemptions can hardly be called full regulation. Europe remains dependent on American AI models from OpenAI, Anthropic, and Microsoft, with Mistral seen as the region's main hope for models on par with the US and China.