EU and Parliament stall on AI Act, deal pushed to next month
▼ Summary
– EU lawmakers failed to reach a deal on amendments to the AI Act after 12 hours of negotiations; talks will resume in May.
– The core dispute is whether high-risk AI systems in products like medical devices and cars should be exempt from the AI Act’s rules.
– Critics, including over 40 civil society groups, argue the proposed changes weaken fundamental rights protections and are a rollback of regulations.
– The Omnibus aims to postpone the AI Act’s compliance deadline for high-risk systems from August 2026 to 2027 or 2028.
– If no agreement is reached before June, the original 2026 deadline stands, leaving unprepared companies facing immediate compliance obligations.
After more than 12 hours of intense negotiations on Tuesday, EU member states and the European Parliament failed to reach a breakthrough on the bloc’s AI Act amendments, pushing a final agreement into next month. The collapse of this trilogue meeting highlights a fundamental rift over whether high-risk AI systems embedded in everyday consumer products should be exempt from the world’s strictest artificial intelligence rules.
“It was not possible to reach an agreement with the European Parliament,” a Cypriot official stated, as Cyprus currently holds the rotating EU Council presidency. This stalled session was the last scheduled political trilogue on the AI Omnibus, a package of changes to the AI Act that took effect in August 2024, alongside proposed tweaks to the GDPR, the e-Privacy Directive, and the Data Act. Framed as a competitiveness measure, the Omnibus aims to slash regulatory burdens so European companies can better compete with US and Asian rivals. However, critics , including a broad coalition of privacy and civil rights groups , contend it is a rollback of hard-won protections disguised as simplification.
At the heart of Tuesday’s deadlock is a single, contentious question: Should high-risk AI systems integrated into products already governed by EU product safety laws , such as medical devices, toys, connected cars, and industrial machinery , be exempt from the AI Act’s additional requirements? The European Parliament, backed by industry groups, argues these systems should only face their existing sectoral rules. The Council, representing member states, has shown little appetite for such a sweeping carve-out.
The Omnibus has drawn sharp criticism from researchers and civil society organizations, who warn that weakening the AI Act before its core provisions have even come into force risks dismantling one of Europe’s most distinctive regulatory assets. Michael McNamara, the Parliament’s lead negotiator on the AI Omnibus, acknowledged in an interview with Tech Policy Press that overlapping rules can be tricky to manage. But he cautioned that shifting AI governance into sectoral laws could ultimately become “deregulatory rather than simplifying.”
Civil society voices have been even more blunt. In mid-April, over 40 organizations signed a letter to the Parliament arguing that the proposed changes undermine the AI Act’s fundamental rights protections, especially for biometric identification systems, AI used in schools, and medical AI. When the AI Act entered into force, it was widely hailed as a global standard-setter.
The clock adds urgency. The AI Act’s core obligations for high-risk AI systems are currently set to apply from August 2, 2026 , just three months away. The entire purpose of the AI Omnibus is to delay that deadline to December 2, 2027, for stand-alone high-risk systems, and to August 2, 2028, for those embedded in regulated products. For the postponement to take legal effect before August, a final political agreement, formal Parliament vote, Council endorsement, and publication in the Official Journal must all happen within weeks.
If talks stall again in May and no deal is reached before June, the original August 2026 deadline will stand. That would mean companies relying on the Omnibus’s extended timelines would suddenly face immediate compliance obligations for which many have not adequately prepared , a scenario Brussels has been desperate to avoid.
The Omnibus does contain one widely supported measure: a ban on AI systems generating non-consensual intimate images, including child sexual abuse material. This was added after the controversy over Elon Musk’s Grok chatbot’s nudification capabilities in late 2025, and both the Parliament and Council had already aligned on it. That talks collapsed despite this area of consensus underscores just how intractable the sectoral exemption question remains.
The resumption of talks next month will determine whether the EU can still claim to be managing this transition in an orderly way, or whether the world’s most ambitious AI regulation stumbles at the very moment its hardest rules are meant to take effect.
(Source: The Next Web)
