PARIS – The global Artificial Intelligence Action Summit is opening its doors in Paris tomorrow, bringing 5,000 AI experts and 80 world leaders to the capital to discuss the latest developments in the world’s hottest technology. 

They hope to agree on a number of international protocols, unveil large investments, and adopt a joint declaration defying sustainable and responsible AI, according to a series of documents obtained by Euractiv.

Imagine the Paris Olympic Games, but AI-style – and with more geopolitics.

The Summit comes as AI technology growth is more unhinged than ever, permeating routines and industrial processes and begging serious questions over national security risks in a field predominantly dominated by the US and China.

Chinese start-up DeepSeek’s meteoric rise in January, seemingly outpacing US competition – in both computing power and cost efficiency – has only intensified the tech race between Beijing and Washington.

But AI cannot be “a conversation solely between the US and China,” an Elysée advisor told Euractiv under the condition of anonymity. “If France and the EU fail to master this technology, we don’t exist.”

The summit, in many ways, is France and Europe’s unspoken bid for relevance in AI governance, funding, and safety.

France’s big bet

Paris already has something to show for it.

The UAE announced last Thursday they would invest some €30 to €50 billion into a large data centre in the country, capable of reaching up to one gigawatt’s worth of capacity. That’s equivalent to powering about a million homes.

France has always trumpeted its cheap, controllable, and decarbonised nuclear energy as a selling point for AI investors.

Expect more blows and whistles as Paris is set to unveil its new AI Foundation for Public Interest to best funnel public financing and provide flagship initiatives with dedicated funding, according to documents seen by Euractiv.

The Foundation will be funded 50% by governments, 25% by industry, and the last 25% by philanthropies, although it is not yet clear how much cash it will be granted – just like it remains unknown how much will be announced through public and private investments.

One thing is certain, however: it will be nothing compared to the US’ $500 billion investment into the Stargate AI infrastructure project.

Who’s in the room?

This is the third global AI summit, after London in November 2023 and Seoul in May 2024.

“Paris will be home to all researchers, thinkers, financiers, investors, and business leaders working on AI issues,” the Elysée touted.

Heavyweights will attend, including Anthropic’s Dario Amodei, Open AI’s Sam Altman, and Google DeepMind’s Demis Hassabis.

Co-hosts Emmanuel Macron and India’s Narendra Modi will take the floor, while political leaders from 80 different countries and EU’s Ursula von der Leyen are expected to show.

Leaders are also expected to agree on a joint declaration on ethical and sustainable AI, enshrining “the need for inclusive multistakeholder dialogues” and “reinforcing the diversity of the AI ecosystem,” according to a draft copy obtained by Euractiv.

AI, the draft reads, must be “human rights-based, human-centric, ethical, safe, secure, and trustworthy” and recognise environmental risks, as a boom in data centres requires vast amounts of electricity and water, 365 days a year.

In effect, the wording of the declaration has been watered down to its most consensual elements. Even then, it may fail to get US backing, as the Trump administration is busy dismantling Diversity, Equity and Inclusion projects and overseeing a significant cutting of red tape.

“We have been incentivised to have a declaration, but it’s not something important,” a civil society lobbyist’s negotiation meeting notes, shared with Euractiv, read.

Investments and concrete actions will tell a more interesting story than the content of the declaration, they explained.

A number of other initiatives should be brought forward, including texts on child protection and copyright laws for artists and journalists.

There will be a call for a strengthened Global Partnership on AI, an OECD-led forum for cooperation that seeks to share best national practices without becoming yet another regulatory authority.

“It’s not another regulatory body,” the Elysée confirmed, fearing the EU may be seen once again as too bogged down on regulation at the expense of innovation. “We shouldn’t overregulate… We need to pivot our approach.”

Jacob Wulff Wold contributed to reporting.

[MM]