This week at the global A.I. summit in Paris, the third iteration of its kind, something noticeably shifted from its previous-year editions (held in Korea last year and the U.K. in 2023). The name, for one. While the inaugural meeting, held in the U.K.’s Bletchley Park, was labelled the “AI Safety Summit,” the 2025 edition changed its title to “AI Action Summit”—a tweak indicative of shifting attitudes amongst its attendees, who were far more interested in A.I.’s opportunities instead of its risks.
“I’ve been to an earlier version of this at Bletchley Park, and that one was more focused on safety,” OpenAI CEO Sam Altman told Bloomberg TV during the summit. “People are now saying ‘Okay, this technology is here, it’s having incredible impact—we’ve got to drive it.”
Economic potential was instead the dominant theme, according to Altman. He said many of his conversations with attendees centered upon boosting investments for A.I. infrastructure in their countries.
“Plug, baby, plug”
Even European policymakers, who have traditionally taken a more conservative approach to A.I. regulation than their American counterparts, appear to be shedding reservations about A.I. “I have a good friend on the other side of the ocean that says, “Drill, baby, drill,’” said France’s President Emmanuel Macron at the summit, in reference to President Donald Trump’s comment during his inauguration speech in January. “Here there is no need to drill. It’s ‘Plug, baby, plug.’”
Macron, who during the conference announced plans to invest more than 100 billion Euros ($104 billion) into France’s A.I. sector, said Europe needed to pick up the pace. “We are committed to go faster and faster,” he told attendees.
This news was well received by U.S. Vice President J.D. Vance. “I like to see that deregulatory flavor making its way into a lot of conversations this conference,” said Vance during a speech at the summit where he criticized “excessive regulation” that could “kill a transformative industry just as it’s taking off.” European countries in particular must “look at this new frontier with optimism, rather than trepidation,” added Vance.
In 2023, the AI Safety Summit resulted in a declaration calling for countries to identify the burgeoning technology’s risks and formulate policies addressing them. Last year’s conference in Korea produced a similar statement. Safety references were significantly scaled back in Paris this year. Instead, the summit’s declaration detailed vague goals to promote A.I. accessibility and ensure the technology is “open and inclusive.” Backed by more than 60 nations, it wasn’t signed by the U.S. and U.K., which endorsed the two previous documents.
Safety-focused A.I. leaders express frustration
Not everyone in the tech industry is on board with the global attitude shift towards A.I. Dario Amodei, CEO of OpenAI rival Anthropic, expressed his frustrations with the AI Action Summit in a blog post. Issues like A.I.’s security risks, its potential misuse by authoritarian countries and its ability to disrupt the global labor market should have topped the conference’s agenda, according to Amodei.
“Greater focus and urgency is needed on several topics given the pace at which the technology is progressing,” wrote Amodei, who described the summit as a “missed opportunity.” Founded by former OpenAI engineers in 2021, Anthropic has long positioned itself as a more safety-focused alternative to OpenAI.
Yoshua Bengio, a renowned expert in machine learning who has been public about his concerns over the technology’s existential threats, appears to share Amodei’s sentiments. More attention must be paid to the risks associated with A.I.’s rapid development, said Bengio in a post on X. “Science shows that A.I. poses major risks in a time horizon that requires world leaders to take them more seriously,” he said. “The Summit missed this opportunity.”
<