The Paris AI Action Summit got under way today with high and broad ambitions. But for whom? A new global foundation is to be unveiled, focused on broadening the AI "ecosystem." So is a new mechanism for protecting intellectual property. China looks set for a warm welcome, even as DeepSeek has shaken the global AI tech order. Safety may remain important, but it is again less prominent. One of the dominant narratives, though, appears to be the desire of the EU, and even more so for France, to secure investment in their AI sectors.
The two-day Paris AI Action Summit is under way with more than a hundred countries represented at a high level, and 1,500 leaders from business, civil society and academia converging on the French capital. At the heart of the summit are discussions today spanning the future of work, sustainability, cybersecurity, collaborative codes of practice and the risks of frontier AI. Then tomorrow, the crowds head to Paris's main venue for startups or to dozens of events around the city, while heads of state meet behind closed doors.
French President Emmanuel Macron is wining and dining Big Tech — and European tech leaders separately — plus leaders including Canadian Prime Minister Justin Trudeau, US Vice President J.D. Vance and, most notably, Chinese Vice Premier Zhang Guoqing. Although India is co-hosting the event and Prime Minister Nerendra Modi is attending, he’s not on the list for Élysée hospitality.
This is the third in a summit series that began in November 2023 at Bletchley Park in the UK. That first edition was dedicated to safety, but it lined up successors: the smaller, more bookish AI Seoul Summit last May, and now Paris.
As well as covering five thematic areas, this week's summit is tasked with ensuring continuity, given that it is seen as unique for bringing government and business together. This seems assured, as Rwandan capital Kigali has already been named in summit documents and a draft communique (see here).
— Global goods —
A new entity is due to emerging from the summit to diversify local AI sectors and make available components for deploying the technology in ways akin to India’s "govstack" modules of digital public infrastructure.
The French title translates as the Global Partnership on AI, though French government officials discussing it at a briefing ahead of the summit referred to it as a “foundation.” Elsewhere it is described as Public Interest AI.
It will be a fund to invest in AI infrastructure that will “irrigate and diversify the AI ecosystem,” Ombeline Gras, global affairs adviser in Macron's private office, told a briefing. It will allow users to develop their own AI with “strategic independence.” India has been helping the development of the modules and databases, in what sounds like a publicly-available toolkit for deploying AI.
It will be based in Paris and will start with 60 nations that will collaborate to avoid “scattered standards,” Gras said. “We obviously need this for our ecosystems, so that the companies we create in Europe can work in the US and vice versa,” she said, adding that it will be government-led, not big tech-led.
Also to be unveiled is the Coalition for Sustainable AI, a "collaborative hub" to highlight how AI can serve environmental goals at a global level. There will also be a closing statement reiterating governments’ commitment to sustainable AI and shared goals.
But there is also going to be a strategy for closer to home. Claire Vernet-Garnier, French presidential adviser on industry and new technology, told a media briefing that a “European AI plan will also be unveiled at the summit, alongside what we are pushing at national level.”
— Intellectual property —
A coalition will be launched to protect the creative sector globally, Gras noted. Ahead of the summit, a document was released outlining a new charter on rights protections, signed by dozens of publishing organizations (see here).
Gras said the World Intellectual Property Organization will also be proposing elements of governance for the protection of intellectual property.
There will also be a new technical solution specifically for news production: “a third party that will enable the media to hand over their data, so that it can be collected and protected by an independent third party who can then monetize it or share it as open source with public research players,” said Vernet-Garnier.
— AI Safety —
Some see the overall scope of the summit series as having expanded so much that safety has naturally become a smaller part of the whole, others see safety as sidelined, with few appearances in the schedule.
But important updates to Seoul’s Frontier AI Safety Commitments are expected. The new International AI Safety Report 2025, commissioned at Bletchley Park and involving the collaboration of 96 scientists, has just been published (see here) and will be presented at the summit.
There should also be reporting from the International Network of AI Safety Institutes, the members of which have been collaborating on model testing in the run-up to the event. The balance of power and focus could shift if the US's institute is significantly changed under Trump, just as China launches its own.
— China —
One narrative strand in Paris is the high profile of Chinese attendees. While Bletchley Park in 2023 attracted Vice Minister for Science and Technology Wu Zhaohui, who signed its declaration, high-level representation was missing in Seoul. In Paris, Macron is not only meeting Vice Premier Zhang Guoqing one on one today, but fellow Vice Premier Ding Xuexiang is also reported to be attending.
Seoul achieved a commitment on frontier AI, signed by 16 tech companies which included one Chinese firm, Zhipu.ai. Just last week, three more were added to the list (see here).
Alibaba’s chief executive is in town, and the French government said it spoke to DeepSeek boss Liang Wenfeng about attending.
— Investment, France and Europe —
"You need to trust AI if you want to have demand for AI, and demand for AI will also determine incentives to continue to invest," the head of the European AI Office, Lucilla Sioli told an event* yesterday.
Her office oversees the EU's AI Act, which is undergoing implementation. Over a thousand organizations including civil society are "participating" in developing the act's codes of practice, said Sioli, in a "co-regulatory and participatory exercise" due to take until May, with a view to finding safe ways to implement frontier AI models.
"After that, we will see how many companies will be willing to sign up," Sioli said, adding that she is expecting to see "constructive participation" from big companies in the exercise.
French government officials were at pains to point out the summit’s potential for investment in Europe and France.
“Around this summit, we are expecting investment in France and Europe, particularly from European players, but also from players all over the world, to show that France and Europe are leading countries in this field,” said Matthieu Landon, industry, research and innovation adviser to Macron.
Backing up his rhetoric, just last night Macron announced 109 billion euros ($112 billion) of domestic and foreign investment in AI in France. More details will emerge today.
Summit organizers, right up to Macron, have explained the role of the EU in organizing the event, albeit with more of a nudge for France. In his foreword to the summit, Macron wrote: “To ensure our continent and our country, which has a rich and vibrant AI ecosystem, are leaders during this century, we need to support our talents and our businesses.”
The EU has also said there will be a European coalition established of companies that will “make the leap and prepare the groundwork to make Europe an incubator and a general hotbed for the growth of future innovative leading actors, which will become tomorrow’s international artificial intelligence champions.”
The EU-level announcements to come from the summit are not yet clear, while among the many global initiatives expected, France’s message of seizing investment while also becoming the guardian of a global public-interest foundation have cut through.
Much work and debate is needed on standards, open-source systems and open access. But the event looks set to serve France and Europe first. In Vernet-Garnier's words: “At this summit, we're not theorizing about AI, we're showing that AI is here, that businesses need to adopt it, that it's a driver of competitiveness for France and Europe.”
*‘Governing in the Age of AI,’ Tony Blair Institute for Global Change, Paris, Feb. 9, 2025
Please e-mail editors@mlex.com to contact the editorial staff regarding this story, or to submit the names of lawyers and advisers.