Mistral CEO Says U.S. AI Safety Debate Masks Competitors’ Negligence
Arthur Mensch called for systems that can contain autonomous AI agents and said Mistral’s next-generation model will close the gap with U.S. labs very significantly.
Mistral CEO Arthur Mensch said the U.S. debate over artificial intelligence safety has obscured what he called competitors’ negligence, while urging companies to build systems that can contain autonomous AI agents.
“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 argued that AI companies need systems capable of containing agents as they gain access to more tools.
“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,” Mensch said. “You need to have the right monitoring in place.”
Mistral provides monitoring systems to enterprise customers, Mensch said.
Anthropic has described three cases in which its Claude models accessed the internet and reached the real systems of three different organizations without authorization during an evaluation. The incidents have added to concerns over how autonomous AI systems behave when connected to external tools and services.
Mistral has no plans to slow development of advanced models. Mensch said the lead held by U.S. AI labs is “not extremely large” and that Mistral expects its next-generation model to close the gap “very significantly.”
The company develops custom AI tools for businesses and integrates them into their operations. It is positioning itself as a European alternative to OpenAI and Anthropic as competition over advanced AI systems intensifies.
Mistral raised $3.5 billion in fresh funding led by Samsung earlier this month. Mensch said the company has been raising capital to expand the computing capacity needed to train larger and more powerful models.
