The Paris AI summit marks a tipping point on the technology’s safety and sustainability

United States Vice President JD Vance made headlines this week by refusing to sign a declaration at a global summit in Paris on artificial intelligence.

In his first appearance on the world stage, Vance made clear that the U.S. wouldn’t be playing ball. The Donald Trump administration believes that “excessive regulation of the AI sector could kill a transformative industry just as it’s taking off,” he said. “We’ll make every effort to encourage pro-growth AI policies.”

His remarks confirmed a widespread fear that Trump’s return to the White House will signal a sharp turn in tech policy. American tech companies and their billionaire owners will now be shielded from effective oversight.

But upon a closer look, events this week point to signs that just the opposite may be unfolding. A host of nations took notable steps towards address growing safety and environmental concerns about AI, indicating that a regulatory tipping point has been reached.

Wide consensus

The two-day global summit in Paris, chaired by France and India, led to broad consensus. Some 60 countries signed on to a Statement on Inclusive and Sustainable AI. This included Canada, the European Commission, India and China.

Both the U.S. and the United Kingdom declined to sign on. But the prevailing winds are against them.

The meeting in Paris was the third global summit on AI, following meet-ups at Bletchley Park in the U.K. in 2023 and in Seoul, South Korea, in 2024. Each of them ended with similar declarations widely endorsed.

The Paris communiqué calls for an “inclusive approach” to AI, seeking to “narrow inequalities” in AI capabilities among countries. It encourages “avoiding market concentration” and affirms the need for openness and transparency in building and sharing technology and expertise.

The document is not binding. It does little more than tout principles, or affirm a collective sentiment among the parties. One of these — perhaps the most important — is to keep talking, meeting and working together on the common concerns that AI raises.

Environmental challenges

Meanwhile, a smaller group of countries at the Paris summit, along with 37 tech companies, agreed to form a Coalition for Sustainable AI — setting out a series of goals and deliverables.

While nothing is binding on the parties, the goals are notably specific. They include coming up with standards for measuring AI’s environmental impact and more effective ways for companies to report on the impact. Parties also aim to “optimize algorithms to reduce computational complexity and minimize data usage.”

Even if most of this turns out to be merely aspirational, it’s important that the coalition offers a platform for collaboration on these initiatives. At the very least, it signals a likelihood that sustainability will be at the forefront of debate about AI moving forward.

Read more: AI is bad for the environment, and the problem is bigger than energy consumption

Signing the first international treaty on AI

A further notable event at the summit was that Canada signed the Council of Europe’s Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law. In recent months, 12 other countries had signed, including the U.S. (under former president Joe Biden), the U.K., Israel and the European Union.

The convention commits parties to pass domestic laws on AI that deal with privacy, bias and discrimination, safety, transparency and environmental sustainability.

The treaty has been criticized for containing no more than “broad affirmations” and imposing few clear obligations. But it does show that countries are committed to passing law to ensure that AI development unfolds within boundaries — and they’re eager to see more countries do the same.

If Canada were to ratify the treaty, Parliament would likely revive Bill C-27, which contained the AI and Data Act.

Read more: The federal government's proposed AI legislation misses the mark on protecting Canadians

The act aimed to do much of what Canada agrees to do under the convention: impose greater oversight of the development and use of AI. This includes transparency and disclosure requirements on AI companies, and stiff penalties for failure to comply.

What does this really mean?

While the U.S. signed the convention on AI and human rights, democracy and rule of law in the fall of 2024, it likely won’t be implemented by a Republican Congress. The same might happen in Canada under a Conservative government led by Pierre Poilievre. He could also decide not to fulfil commitments made under other agreements about AI.

And if Poilievre comes to power by the time Canada hosts the next G7 meeting in June, he might decline to honour the Trudeau government’s commitment to make AI regulation a central focus of the meeting.

The Trump administration may have ushered in a period of more lax tech regulation in the U.S., and Silicon Valley is indeed a key player in tech — especially AI. But it’s a wide world, with many other important players in this space, including China, Europe and Canada.

The events in Paris have revealed a strong interest among nations around the globe to regulate AI, and specifically to foster ideas about inclusion and sustainability. If the Paris summit was any indication, the hope of sheltering AI from effective regulation won’t last long.

Robert Diab does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.