APREX · EST. 2026
The EU's Artificial Intelligence Act is facing delays once again. The European Parliament has recently voted to postpone a number of key deadlines in the regulation — a decision that many industry players are describing as a significant victory following months of intense lobbying.
Which Deadlines Have Been Postponed?
The specific changes are wide-ranging. According to The National Law Review, the compliance deadline for high-risk AI systems used in employment, education, and law enforcement — governed by Annex III — has been pushed from 2 August 2026 to 2 December 2027. That represents a delay of 16 months.
For AI embedded in regulated products such as medical devices and industrial machinery (Annex I), the deadline moves from 2 August 2027 to 2 August 2028 — a 12-month extension.
Requirements for labelling AI-generated content have also been revised. Providers of systems already on the market before August 2026 now have until 2 December 2026 to ensure that synthetic content is machine-readable and identifiable as AI-produced. In addition, EU member states have until August 2027 to establish national regulatory sandboxes for AI.
The Industry Pressure Behind the Delays
The delays are no accident. According to research materials, the changes are a direct result of sustained pressure from European businesses, combined with concerns about the EU's ability to deliver the necessary guidance documents and technical standards on time.
One of the core problems is that European standardisation bodies have fallen behind schedule. Many of the harmonised technical standards that companies need in order to understand their obligations are now not expected until late 2026 at the earliest. Without these tools, businesses would have faced what is described as "significant uncertainty" around compliance.
More than 45 leading European companies publicly called on the EU to pause the rollout of high-risk AI rules, pointing to "unclear, overlapping, and increasingly complex" regulations
In a letter dated 3 July, more than 45 prominent European companies signed a public appeal urging the EU to pause the implementation of rules for both high-risk AI and general-purpose AI. They argued that the regulation as it stood threatened to upset the balance between oversight and innovation.
More than 30 founders and venture capital executives submitted their own open letter, warning that an incomplete framework would disproportionately harm startups — and in practice favour large, established corporations with the resources to absorb high compliance costs.
Commission: 'This Is Not a Watering-Down Exercise'
In November 2025, the European Commission published the so-called "AI Omnibus" document, introducing measures to simplify and enable a more proportionate implementation of the law. EU officials stress that the changes relate solely to timing, not to reducing the regulation's level of ambition.
Critics, however, remain unconvinced. Analysis by Corporate Europe Observatory and LobbyControl shows that 69 percent of the Commission's meetings in 2025 were held with business groups — which, according to these organisations, illustrates the industry's considerable influence over how the changes were shaped.
Overlapping Regulations Resolved
The changes also resolve one specific issue that has long caused headaches for manufacturers: AI-enabled machinery and industrial equipment previously had to comply with both the AI Act and existing sector-specific safety legislation simultaneously. This has now been clarified — such products will primarily be governed by the relevant sectoral legislation.
The question going forward is whether the delays will give industry the predictability it is seeking — or whether they will in practice open the door to further pressure for additional postponements as the new deadlines draw near.