How Emmanuel Macron brought France’s AI darling to the world stage

French President Emmanuel Macron personally championed AI startup Mistral, helping it secure a partnership with Nvidia and interest from BNP Paribas, Orange, SNCF, and Thales, according to POLITICO interviews. Mistral announced a €3 billion funding round on Tuesday, valuing it at over €21 billion, making it Europe's leading AI player, but its future is uncertain as Macron's presidency ends.

PARIS — It’s not every AI company that can count the French president as its top salesman. On June 11, 2025, the founder of French AI company Mistral took to the stage alongside Jensen Huang, the powerful CEO of chip giant Nvidia, to announce a partnership between the two companies. As President Emmanuel Macron joined them, the Taiwanese-American executive turned to Mistral’s young CEO Arthur Mensch and said something just audible to the Parisian audience of techies and public officials: “Your president worked really hard for you.” The Nvidia boss was referring to the deal his company had just sealed with Mistral to roll out AI infrastructure across Europe. “I told the president that Arthur and Mistral needed support from France’s biggest companies. Mr. President said to me: ‘Which ones? Let me call them ’” Huang said. Advertisement By the time the deal was official, the companies could announce that BNP Paribas, Orange, SNCF and Thales had all expressed interest in the partnership — boosting Mistral’s official entry into the global race for computing power. Even for the head of Nvidia, the world’s most valuable company, such presidential proactivity was remarkable, illustrating the French government’s dedication to championing its homegrown startup and showing how Mistral has leveraged ties to the executive branch in Paris to try and reach the level of some of the biggest generative AI companies. Mistral’s rise over the past three years, according to a dozen public officials and representatives of France’s tech ecosystem interviewed by POLITICO, owes much to the privileged relationship it built with Macron — enabling the company to secure both strong ties with the French state and a seat at the table with the world’s leading AI companies. Now, with a French presidential election bringing to a close Macron’s two-term reign, details on the depth of that relationship — many of which are reported here for the first time — also raise questions about Mistral’s future under a different president. The company on Tuesday announced a new €3 billion funding round including investments from the European Union and South Korea’s electronics giant Samsung, valuing Mistral at more than €21 billion. While that falls well short of the valuations of Anthropic and OpenAI, the news confirms the French startup as the only serious European player in the AI race. As Europe questions its ability to challenge Chinese and American tech giants, Mensch and his team must now plan for life after Macron. Advertisement ‘Fairy godmothers around the cradle’ Launched in 2023 by three co-founders — Mensch, Guillaume Lample and Timothée Lacroix — Mistral AI set out to build a full-fledged large-language model to compete with the likes of OpenAI and Anthropic. To do so, the Paris-based startup put forward a clear strategy from day one: cultivate a relationship with the French executive, and Macron in particular. From its first funding round, the startup — which declined to respond to questions for this article — made a point of forging ties with the public sector. First among them was Bpifrance, France’s public investment bank and one of its earliest investors. “We were aware of the sovereignty issues linked to AI,” recalls Arnaud Caudoux, deputy CEO of Bpifrance. “We had a good idea of the few — a few hundred in the world at the time — credible experts in the sector, among whom were Mistral’s founders.” The BPI investment was part of a record €105 million total for a French seed round. But Caudoux noted: “There were a lot of fairy godmothers around the cradle.” Those looking out for the startup came from the cream of French business and tech, playing crucial roles as matchmakers, patrons and fundraisers. They included telecom billionaire Xavier Niel, as well as Jean-Charles Samuelian-Werve and Charles Gorintin, co-founders of insurance startup Alan. They were driven by the same goal: helping a European AI giant emerge. They tapped their networks to identify the most promising French AI researchers. These three businessmen were the ones who played matchmaker, bringing together Mensch, Lacroix and Lample. Advertisement When the Alan team struck a deal with the three young co-founders, it secured a strategic recruit to forge ties with the public sector: Cédric O, a former digital affairs secretary of state and early Macron supporter, would play a crucial role in the startup’s launch. It was O who went into battle in Brussels to try to secure looser rules under the European Union’s AI Act, first proposed in April 2021. For Mistral, the burgeoning legislation posed a threat: How could Europe build a giant of its own if it stifled innovation? In 2023, O intervened on Mistral’s behalf in EU talks in an effort to influence the negotiations — to the point that some lawmakers accused him of a conflict of interest. He also drew a rebuke from Thierry Breton, then European commissioner for the internal market. Mistral “is lobbying, that’s normal,” French-born Breton said during the final weeks of negotiations over the AI Act. “But we are not fooled by anything. It is defending its business today, and not the general interest.” “Cédric O’s role was, in a way, that of a lightning rod,” said a private-sector official who followed Mistral in its early days. By taking the hits, O allowed Mensch “to grow and develop as a CEO.” Behind the scenes From the outset, Mistral played the French and European card. The three co-founders may have built their careers in U.S.-based AI labs — Google DeepMind for Mensch and Facebook AI for Lample and Lacroix — but the first concept notes drafted for the startup’s launch put technological sovereignty at the heart of the project. Advertisement At the time, Mensch was a discreet 30-year-old engineer who needed to transform himself into the public face of Mistral on the international stage. Just a few months after launching the startup, the alum of the prestigious École Polytechnique made his first public appearance at VivaTech 2023 — Europe’s largest annual technology event — alongside Macron, who had already begun making the appeal for European technological autonomy. The two were clearly in sync. “How do France and Europe intend to enable European champions to emerge, and how do we make sure that AI regulation does not hinder innovation?” the young sneaker-wearing entrepreneur asked the French president on the main stage. Mensch would himself be tasked with helping come up with the answers. A few months later, the prime minister’s office formed a committee of experts to devise a French AI strategy. Alongside Mensch and O, its members included Gaël Varoquaux, a top researcher at France’s national digital research institute and Mensch’s former Ph.D. supervisor, as well as future Nobel economics laureate Philippe Aghion and Meta researcher and Turing Award recipient Yann LeCun. By 2024, the committee delivered its report and Mistral unveiled Le Chat, its rival to OpenAI’s ChatGPT. But the AI landscape was already changing rapidly — and Europe was having second thoughts. German rival Aleph Alpha, the continent’s other great hope in generative AI, announced it was moving away from developing its own models. Its CEO, Jonas Andrulis, offered a gloomy assessment: “Simply having a European LLM is not enough as a business model. It does not justify the investment.” Mistral was not spared those doubts. Many observers feared its technology could fall behind American or Chinese rivals — even raising concerns that the startup might be acquired from abroad. Advertisement In 2025, reports of talks with Apple sparked panic at the highest levels of the French state. France’s embassy in the United States and the Elysée presidential palace asked Mistral whether the reports were true, according to two advisers who spoke to POLITICO. The company eventually reassured them: There had been no discussions “at the highest level.” The French champion instead turned to a partner closer to home, and at the heart of Europe’s drive for technological sovereignty: the Netherlands’ microchip-printing giant ASML. The Dutch pioneer in the chips supply chain poured €1.3 billion of fresh investment into Mistral in September 2025 in a bid to speed up industrial AI development and keep Mistral’s know-how in European hands. The deal was a boon for politicians — including Macron — seeking to keep the United States out of Europe’s most prized critical technology players. Strategic partner Three years into the company’s existence, the French government is no longer simply promoting a promising young startup. The state is now prepared to bend over backward to preserve a company whose technologies are being integrated across government. After signing a strategic partnership with France’s national employment agency in early 2025, Mistral secured a deal with the French Armed Forces Ministry to develop AI for defense. Since 2025, Le Chat has also been in the process of being rolled out to 10,000 civil servants. In public, Mensch has played down the importance of public subsidies and government support for his company. But Mistral’s teams in private conversations readily acknowledge that this privileged relationship is a major asset. Advertisement With Macron’s help, the French company has also opened the doors to the world’s biggest international summits — all the way up to the G7. Mensch has been brought along on almost every presidential trip abroad in the past two years, giving him repeated opportunities to sharpen and spread his AI counter-narrative to contrast with the all-powerful Silicon Valley vision. For several months, the Mistral boss — who readily appropriates the École Polytechnique motto, “For the Nation, Science and Glory” — has hammered home a position aligned with Macron’s: Europe must promote technological sovereignty and offer the rest of the world a model that is more sensitive to cultural and linguistic nuances than American and Chinese rivals. In February, on stage at the New Delhi AI summit attended by Dario Amodei of Anthropic, Demis Hassabis of Google DeepMind and Sam Altman of OpenAI, Mensch defended AI as “a tool for empowerment, not domination.” He spoke about the importance of deciding who controls the deployment of AI, who benefits from it and how to ensure that it serves the general interest, not just a few powerful players. That landed as countries across the world were scrambling to get insights from U.S. firms and the Trump administration into how new AI models were creating major new hacking risks. And in June, in Evian-les-Bains at Lake Geneva, Mensch had lunch with G7 leaders and AI executives, including Altman. He once again took on the role of maverick to confront his American rivals, who were calling for technological standards to be drawn up under U.S. leadership. Advertisement “We still need to be very wary,” Mensch told POLITICO in an interview after the meeting. “It’s true that American actors tend to make that argument, which is, ‘Trust us—we’ll set standards here, and then we’ll export them to you.’” Succession planning As Mistral moves into infrastructure by building data centers and continues rolling out its solutions at major companies, some are beginning to doubt whether it can remain in the race to produce “frontier” models — the most powerful models available. For Macron, the June G7 meeting was one of the last opportunities to leave his mark on the global AI debate before his presidency ends next spring. For Mistral, its relationship with Macron’s successor has yet to be built. Gabriel Attal, former prime minister and presidential candidate for Macron’s pro-business bloc, readily showcases his ties to Mistral and has put AI at the heart of his campaign platform — even incorporating Le Chat into his communications toolkit. Attal’s centrist rival Edouard Philippe would also almost certainly continue backing Mensch. While Mistral does not publicly associate itself with the far-right National Rally of Marine Le Pen, the nationalist party sees the company as a sovereign asset. On the left, the parties have unanimously backed the company — and called for sovereign AI — in response to the risk of American models being weaponized against European countries. Yet Mistral relies on international funding and major partnerships with American giants including Amazon, Nvidia and Microsoft. While bringing in ASML as a key investor meant Mistral would remain firmly rooted in Europe, it also shifted some of the power to ASML’s headquarters in Veldhoven, the Netherlands. Advertisement This detail could undermine the French consensus in favor of Mistral, just as AI is emerging as a key campaign issue for the presidential election. “We need to kill this very bar room idea that it would be feasible to own the capital, and therefore finance on our own, at French or European level, the hundreds of billions that AI infrastructure requires. It’s simply impossible,” said Bruno Bonnell, a French businessman who runs a major public investment program, France 2030, under the prime minister’s office.