EU officials have agreed to relax some aspects of AI law, including delaying implementation of rules covering a number of high-risk applications until December 2027, instead of the originally set deadline of August 2026, according to Latest update from EU lawmakers Relaxing AI rules.
The agreement comes after several companies claimed that the EU was drowning itself in unnecessary regulation, leaving the EU behind competitors in the US and Asia.
The agreement was reached after 9 hours of talks, which is fairly standard for negotiations in Brussels. It still needs to be ratified by EU leaders and the EU Parliament, so don’t expect any final changes yet. But the conclusion is quite clear: Europe still wants to regulate AI, but less stringently.
The final deal means that high-risk autonomous AI systems must comply by December 2, 2027, but high-risk systems embedded in high-risk products, such as cars or medical devices, will have until August 2, 2028 to get it right.
The Council said this was intended to help “simplify” AI law, including by preventing overlap with other sectoral legislation. In other words, if the machine, medical product or device was already regulated as a regulated product, there would be no need for companies to produce duplicate paperwork just to comply with the AI Act.
However, the deal is not a golden ticket for major AI companies: the agreement will impose a ban on non-consensual sexually explicit images and videos, including so-called “nudity” apps and child sexual abuse material.
The ban is scheduled to come into effect on December 2, 2026, when watermarks on AI-generated content take effect – allowing for a clearer timeline for industry players.
The European Parliament said the simplifications package for AI law “strikes a delicate balance between simplifying the rules and preserving the risk-based approach to AI.” Artificial Intelligence Law and Adding Safeguards Against So-Called “Nudity Apps”“.
It’s a crucial point – few would really argue that we should delay tackling the problem of sexual deepfakes, especially now that women, young people and politicians have made themselves the targets of synthetic images, images that are not only harmful but destructive.
The main disagreement is about timing. Civil society and digital rights activists contend that delaying stricter regulations around high-risk AI means leaving individuals exposed in a variety of fields, from employment and education to biometrics, critical infrastructure and policing.
Conversely, the business community contends that an unclear landscape with overlapping commitments would stall Europe’s AI industry before it really takes off. Either could be true, making this a minefield.
The original law went into effect in August 2024, when It has been hailed by the European Commission as the first full regulatory framework for AI In the world. The law is risk-based: some uses of AI are prohibited, high-risk uses have strict requirements, and low-risk uses have lighter obligations. This remains the case under the new agreement, which only delays the timing and scope of some of the more stringent obligations.
It all sounds like a political hit. Europe has positioned itself for years as the responsible party in the AI conversation: the one that prioritizes rights and safety over hype.
Now, under intense pressure from industry and big technology, it is beginning to decline. Pragmatism? Yes. surrender? You can be sure that many will argue that. I think the truth lies somewhere in the messy gray area in between.
Siemens and ASML have pushed for AI regulations for industrial applications, with Reuters reporting that AI Act rules will not apply where there are industry-specific regulations.
For manufacturers who have been concerned about compliance headaches, especially in some of Europe’s industrial powerhouses, this represents a welcome development. It also raises a simple question: When does simplification become a vulnerability?
the The European Commission praised the agreementnoting that the amended Artificial Intelligence Law aims to encourage innovation while protecting citizens from the harmful consequences of artificial intelligence. “Innovation and protection”, “speed and safety”, “less paperwork, more human rights” – everyone wants it; Nobody wants it to be true.
For startups, the postponement offers some relief. In the European Union, building AI has become a regulatory minefield, and small companies may lack Google’s resources in the form of a team of compliance experts.
If the AI law takes longer to implement, it could give European developers more room to compete rather than spending money on law firms once they start working on seeds.
But the settlement doesn’t look very nice to the public. High-risk AI systems are classified as “high risk” for a reason — they can affect who gets hired, how governments deliver services, how police use their tools, and even how critical infrastructure works. Delaying implementation may reduce industry concerns, but it also delays the day when citizens receive maximum protection. It is a difficult dilemma that Brussels will not be able to overcome.
Europe wants to be the region that sets the rules for the AI era. But it also wants to be the place where AI companies build real-world products. Both of these goals can be achieved, but they must be applied with enough friction to create a little heat. This week’s agreement aims to alleviate some of these frictions before they worsen.
The final compromise will move to the next stage of the formal process and, if approved, will set the course for the first few years of AI law implementation, while also providing a signal to countries outside the EU that even the world’s most ambitious AI regulators are adjusting their plans based on the pace, costs and political realities of the AI race.
Now, the real question is: Does Europe still want to impose strong AI rules? Obviously it does. But can Europe also make it enforceable while not making it so weak that the safety shield leaks?
