A French AI giant championed by Emmanuel Macron, the French president, is on the hunt for British defence work as it seeks to sell itself as an alternative to Silicon Valley.
The €21bn (£19bn) Paris-based lab is hiring experts who can help it break into deals working on UK security and critical infrastructure.
The French business, which develops AI technology similar to that of OpenAI’s ChatGPT and Anthropic’s Claude, has previously warned that Europe risks becoming over-reliant on US technology.
Arthur Mensch, its chief executive, said earlier this year that Europe would become a “vassal state” of the US if it failed to build its own technology.
Mistral has launched AI chatbots, coding tools and autonomous digital “agents”. Last month it launched its most powerful new bot aimed at coding and cybersecurity, nicknamed “Le Chonk”.
Its AI is designed to be “open weight”, similar to the technology that is pouring out of Chinese AI labs. This means that developers can adjust and tinker with the AI tool freely, often at a lower cost.
Mistral has already been looking for defence tie-ups in France as it pitches its AI as a system designed and built for Europe after the US imposed restrictions on exports of some of Anthropic’s most powerful systems.
A White House decision to block Anthropic’s most powerful AI, Mythos, from widespread access has led to fears the US could cut off Europe from powerful forms of superintelligence.
In January, Mistral signed a deal with the French army to supply its technology for up to three years.
Mr Mensch has also insisted his AI lab would not interfere with how governments or militaries choose to use its bot, after Anthropic executives clashed with the Pentagon earlier this year over safeguards built into its technology.
He told AFP in May this year: “Who are we to tell soldiers, who know their job and the dangers … what they’re allowed to do?”
Mistral is now targeting similar work in the UK. A job listing for an AI deployment strategist specialising in defence calls for expertise working with “sovereign institutions and critical infrastructure across the UK”.
The AI lab also hired a so-called go-to-market executive with experience of national-security companies.
The Government has faced calls to reduce its reliance on US technology businesses, with Labour MPs demanding ministers end a deal with Palantir, the American data giant.
However, officials are now expected to continue with the £330m deal to its conclusion.
Earlier this year, ministers announced plans to develop open-source AI tools aimed at national security, working with Meta, which has focused on developing AI systems that can be freely downloaded.
The Telegraph previously revealed that the Ministry of Justice had been using Qwen, an open AI model from Alibaba, for research analysing criminal records.
Mistral was contacted for comment.
Leave a comment