French AI company Mistral says its newest model outperforms Chinese models on cybersecurity. CEO Arthur Mensch made the claim on Tuesday, October 6, at the Ai Everything conference in Abu Dhabi. He said the company would unveil the model later the same day. Mistral is three years old and is seen as Europe's strongest contender in the AI race. It is trying to compete on performance while also presenting itself as a safer, independent option.
Also read: OpenAI's Sam Altman Says AI Benefits Justify Accepting Some Risks
On stage, Mensch said the new model is "above the Chinese models on certain aspects, including cyber." He argued this proves the idea that Europe cannot compete in AI is not true. He did not say which Chinese models he meant or how Mistral measured the gap. No benchmark results have been shared, so the claim cannot be checked yet. It also covers only specific areas, not overall performance.
Mensch added that Mistral's independence from both the United States and China helps it grow in the Gulf and Asia-Pacific regions. He said many customers want other technology suppliers.
Also read: Latest Developments Around AI Safety & Governance
Mistral raised 3 billion euros at a 21 billion euro valuation in September. Even so, it remains far smaller than U.S. firms such as Anthropic and OpenAI. It also faces pressure from cheaper Chinese rivals. Its last model launch came in April.
Security can decide whether a company wins a government or enterprise deal. A strong cyber result could help Mistral win regulated and public-sector customers in the Gulf and Asia, where trust matters as much as speed. Those deals tend to bring in more revenue than low-priced AI tools that customers can swap out easily.
The real test comes once outside experts run their own tests. Until then, the claim is a bold sales pitch with little proof behind it.