Arthur Mensch

Arthur Mensch is the co-founder and chief executive of Mistral AI, the Paris-based large language model developer that has become Europe's most credible challenger to the American AI giants. Born in Sèvres, France, in 1992, Mensch trained as an engineer at École Polytechnique and Télécom Paris before completing a master's in the Mathematics, Vision, Learning programme at ENS Paris-Saclay and a PhD in machine learning at Inria, supervised by researchers including Bertrand Thirion and Gaël Varoquaux.

Before founding Mistral, Mensch spent roughly two and a half years at DeepMind's Paris office, working on the large language models and multimodal systems that underpinned Google's Gemini effort. That vantage point, inside one of the world's best-resourced AI labs, convinced him and co-founders Guillaume Lample and Timothée Lacroix that Europe needed its own sovereign alternative, one not beholden to Silicon Valley's cloud platforms or Beijing's state apparatus.

Mistral AI launched in May 2023 and moved with unusual speed, closing a record-setting €105 million seed round within weeks and releasing its first open-weight model, Mistral 7B, before the year was out. The company's central bet has been openness: releasing model weights rather than keeping them locked behind an API, a strategy Mensch has defended publicly as both a commercial differentiator and a matter of digital sovereignty for governments wary of dependence on US infrastructure.

Funding has followed the momentum. Mistral raised €385 million in December 2023, €600 million in a Series B in mid-2024, and roughly €1.7 billion in a September 2025 round that pushed its valuation to around €11.7 billion, with Microsoft, ASML and Nvidia among its strategic backers. Under Mensch, the company has expanded from pure research into enterprise products, coding assistants and government partnerships, positioning itself as the anchor of a European AI stack distinct from both US hyperscalers and Chinese labs.

Through 2026, Mensch has continued to press the case for open models and European AI infrastructure in prominent interviews, including discussion of Mistral designing its own chips and data centres to reduce reliance on Nvidia and US cloud providers. His profile has grown alongside the company's, making him the most recognisable spokesperson for the argument that AI's future need not be dictated solely by American and Chinese labs.