EU resets relations with AI industry, but at what price?

Of all the big summits I have reported on over the years, the ones that yield the least are those that carry the most pompous and wide-ranging agendas. The Paris Artificial Intelligence Action Summit, held this week, risks being one such event. It saw heads of state, top government officials, CEOs and tech company executives and scientists, in addition to other market and industry stakeholders and civil society from more than 100 countries, rub shoulders while grappling with the real-world impacts of AI.

Two years on from the emergence of OpenAI’s ChatGPT chatbot, AI has been fueling both immense hope and, at times, exaggerated fear, depending on which side of the equation you stand.

The Paris summit was the third such gathering. The first was held in 2023 in the UK and resulted in a nonbinding pledge by 28 nations to tackle AI risks. A follow-up meeting hosted by South Korea last year secured another pledge to set up a network of public AI safety institutes to advance research and testing.

Participants in Paris reaffirmed the fact that the world is experiencing a tech and scientific revolution the likes of which has rarely been seen before. But they did not offer much in terms of enabling people to live better, learn better, work better or care better. They again failed to bridge the gap between the stakeholders and developers of AI and the market, the consumer and regulators, particularly when geopolitics and state interests enter the fray. Is technological prowess not always synonymous with military and political power?

And, as expected, the world stood divided once more, with the US and UK on one side and the EU, China and India on the other. The Paris meeting’s closing declaration agreed to prioritize regulation to make AI’s development “open” and “ethical,” but neither the UK nor the US, which is home to the world’s largest AI companies, signed the agreement. The Americans claimed that regulation kills AI and the British insisted that the declaration did not sufficiently address “harder questions around national security.”

The organizers tried to get countries to commit to a more ethical, democratic and environmentally sustainable AI. But in the fragmented world in which we live, the time has not yet come to put competition aside in favor of the common good. The recent spat, when a cheaper AI model emerged in the form of China’s DeepSeek, is very telling about how the world is still too far apart to come together and agree on governance, regulations and due diligence to limit the harms caused by AI.

The French, like many in the EU, wanted to use this summit to play catch-up and claim a share of the development and ownership of AI, as it has been concluded that the technology is not mere hype but a game-changer on all levels. And those who fail to tap into it could be left behind politically, economically, socially and even militarily. So, France and Europe in general are looking increasingly like they are in the process of abandoning their guardrails in the race to develop better-than-human AI through sheer computing power and Paris’ initial calls to direct AI toward solving society’s problems, particularly the ills of industry and the health sector, have become muted.

It seems that alignment with the age of Trump 2.0 is high on the agenda of the old continent’s leaders, regardless of what they say. They are increasingly certain that the world has changed and the old norms of nominally respecting the rule of law and multilateral institutions and governance are increasingly irrelevant. The US does not shy away from asserting its desire to become the world capital of AI development through tapping its oil and gas reserves to feed this energy-hungry technology. Also, Donald Trump’s narrative points to a lukewarm approach to even light regulation, as he seeks to maintain America’s global supremacy and leadership through reducing regulatory barriers and building AI systems free of any ideological biases.

The EU, despite being a third entrant in a two-horse race led by Washington and Beijing, could have a shot at making an impact, but only if it reverses years of red tape and regulation aimed at protecting society from the excesses of AI technology and its potential harms.

DeepSeek’s release of its latest model last month stunned the world because of its ability to rival Western players like ChatGPT. It also escalated the wider geopolitical showdown between Beijing and Washington over tech supremacy. But it also showed the EU that there are possibilities for a new breakthrough that could tilt the balance toward Europe, effectively making AI a three-horse race.

DeepSeek showed the EU that there are possibilities for a new breakthrough that could tilt the balance toward Europe.

Mohamed Chebaro

The EU’s messaging around the summit has reflected a clear divergence from its former rhetoric in a bid to soften its narrative, as well as its rulebook. Henna Virkkunen, the European Commissioner for tech sovereignty, security and democracy, agreeing with the tech industry about the need to review the EU’s regulatory regime is unprecedented. Her calls to review overlapping regulations and remove the administrative burden from the EU’s tech industry have been applauded, but the question is how far will this go? Will the EU even do away with its AI Act, the world’s first comprehensive set of rules governing the technology?

If anything, the Paris summit sadly showed that making AI more sustainable for people and planet is out, as more of Europe’s productivity is dependent on being a leader in this emerging technology and using it to serve its member states, businesses and societies, while securing its independence in the tech realm in a way that serves national interests.

For now, it seems that the way forward for the EU is to attempt a lighter-touch approach to AI regulation. The worry is that weakening existing regulations might not be enough, as the world — led by Trump and his tech partners — is driving the agenda and both directly and indirectly pushing us toward more uncertainty, fewer human rights and AI that will ultimately challenge existing norms and human skills, rendering some jobs redundant and all workers less protected.

• Mohamed Chebaro is a British-Lebanese journalist with more than 25 years’ experience covering war, terrorism, defense, current affairs and diplomacy.