BRUSSELS — The European Union’s rules to rein in artificial intelligence risks are barely a year old but they’ve already fallen out of favor.
Fueled by a new political climate and heavy lobbying from industry, an effort to revisit the carefully negotiated law is spurring warnings from safety campaigners and introducing fresh uncertainty that some industry groups say could undermine investment in Europe.
When the bloc’s AI rules were agreed, EU officials and legislators celebrated a multi-year effort to set a global standard designed to keep humanity safe from dangerous tech.
But the focus has shifted from safety to catching up with American and Chinese companies leading the AI race, and the EU’s prized AI rulebook is at strong risk of being tweaked as part of a push in Brussels to simplify laws and reduce requirements on businesses.
While digital rights campaigners warn the change of heart could see technological catastrophe come true, industry is also conflicted about the indecision.
“If you actually want to impose something, impose it, if you want to stop the clock, let’s stop the clock,” said Tomasz Snażyk, CEO of the Polish business group AI Chamber. “People want to be very certain of what is going on.”
Dutch Greens’ lawmaker Kim van Sparrentak, who is an advocate of simply enforcing the laws that she herself helped negotiate, warned about the potential brake on the adoption of technology at a time when only 13 percent of companies in Europe are estimated to properly use AI.
“As long as we don’t have an AI Act that is being enforced, we keep systems on the market of which you don’t know whether these function well or [if] these are safe. That to me is essential if we want to ensure that a lot of companies start using AI,” said van Sparrentak.
That’s in contrast to the Big Tech industry backed by the U.S. government, which argues a delay — at a minimum — is vital to give the industry the time it needs to comply.
“The postponement is essential … to give companies enough time to prepare,” lobby group CCIA said in a recent paper.
‘Deeply concerned’
Big Tech companies in January gained a powerful ally in the White House as Donald Trump’s administration pushed back against the EU’s Artificial Intelligence Act.
Yet it was still a surprise when the European Commission’s tech sovereignty chief Henna Virkkunen this month opened the door to pausing the roll-out of the law.
While the EU rulebook became law in August 2024, it only takes effect gradually: Some AI practices were prohibited as of February, but new laws against high-risk AI applications will only take effect in 2026 or 2027.
That timeline could be put on pause, Virkkunen told EU ministers in Luxembourg on June 6. “If we see that the standards and guidelines … are not ready in time, we should not rule out postponing some parts of the AI Act,” she said.
Adding to concerns is an ongoing Commission review of how to simplify all its digital rules, to be presented in December.
Virkkunen has repeatedly stated that will focus on AI, cybersecurity and data rules. Kilian Gross, a top official inside the Commission’s AI Office, left the door open for “targeted” changes to the AI Act.
That’s prompted dismay from campaigners and lawmakers.
“We are deeply concerned by the ongoing attempts to deregulate key provisions of the AI Act,” said Blue Duangdjai Tiyavorabun, a policy adviser at EDRi, which advocates for digital rights. Europe’s vision of the future shouldn’t be built on “dismantling hard-won legal protections in a matter of months.”
Lawmakers are set to grill Virkkunen on Tuesday in Strasbourg during a meeting of the parliamentary group that monitors AI Act implementation. The possible pause “will dominate” that meeting, said Irish liberal lawmaker Michael McNamara, who co-chairs the meeting.
“Just to throw everything out, there would be a risk of throwing out the baby with the bathwater,” he said.
McNamara urged Virkkunen to present more details to the Parliament group. “What does she intend to postpone and for how long?” he asked.
Final nail
The indecision could be the final nail in the coffin of the so-called Brussels effect: the idea that EU rules on technology are world-leading and have global repercussions.
If Europe succeeds in scaling homegrown AI companies that can rival some of the U.S. frontrunners, that counts as a massive win for the bloc’s industrial policy objective.
But it must do so in a way that doesn’t throw into question its ability to regulate.
The Commission needs “to give a signal that this is our law [and] this is what we stand for,” said van Sparrentak.
Part of the problem is that companies need standards and further guidance to comply with some of the requirements of the law, such as around the transparency of models and keeping humans in the loop — and those guidelines from the Commission have not yet appeared.
That led industry groups that were already against the laws to publicly and visibly call for a postponement. The U.S. government also called for a pause in implementation in feedback sent to the European Commission in April.
Van Sparrentak branded Tuesday’s meeting as an opportunity for Virkkunen to show the Commission is not bound to White House whims: “It’s super unwise to start listening to Trump. This was his appeal: stop the clock,” she said.
The Commission needs to “stand up that we have rules which will ensure that AI systems in Europe are safe,” she said.
European Commission spokesperson Thomas Regnier said that the Commission is still “fully committed to the main goals of the AI Act,” but added that as it seeks to simplify its digital rules “all options remain open for consideration.”
The post EU’s waffle on artificial intelligence law creates huge headache appeared first on Politico.
