Mistral CEO tells CNBC US AI safety debate covers rivals’ ‘negligence’

Mistral chief executive Arthur Mensch said the US debate over AI safety has been a cover for the “negligence” of some of his rivals. He made the comments to Annette Weisbach in an interview with CNBC on Tuesday.

“The debate that we’ve seen in the U.S. has been a cover for the negligence of some of our competitors,” Mensch said.

He said companies need systems that can contain AI agents.

“When you give them a lot of tools, those systems are very dynamic, so they can go and do things that you do not expect,” he said.

Mistral gives the businesses it works with monitoring systems for their agents, Mensch added.

A month of warnings

Mensch spoke a day after OpenAI cancelled the release of GPT-6.1 Astra. The company said the model did not meet its safety standards.

Earlier this month, an OpenAI agent got into an Australian government website without permission. An Anthropic researcher quit the company, saying AI labs were “gambling with our lives”. Anthropic chief executive Dario Amodei then urged labs to slow the pace of improving their models.

In July, Anthropic described three cases in which its Claude models reached the internet during a test. Each time, they got into the real systems of an organisation, CNBC reported.

Mensch is not the only critic of the US debate. David Sacks co-chairs the President’s Council of Advisors on Science and Technology.

“Stop pretending the motivation to slow down is purely altruistic,” Sacks wrote to tech companies on X, as quoted by CNBC.

Emil Michael, the Pentagon’s under secretary for research and engineering, has warned of a “coordinated campaign” of fearmongering.

‘Not extremely large’

Mistral builds custom AI tools for individual companies. It has no plans to slow the development of its advanced models, CNBC reported.

Mensch said the lead of the US labs is “not extremely large”. He expects Mistral’s next-generation model to close the gap “very significantly”.

“We have been raising the capital needed to scale the compute that we need to train bigger and more powerful models, in order for us to own our own destiny,” he said.

Mistral raised €3bn earlier this month, in a round led by Samsung.