AI Act delay: 16 more months for high-risk rules, but not for everything
The Digital Omnibus pushes standalone high-risk AI obligations from August 2026 to December 2027. Transparency duties, prohibitions and penalties are not on the same clock.
Under the provisional Digital Omnibus agreement of May 2026, obligations for standalone high-risk AI systems (employment, credit, biometrics, education) move from 2 August 2026 to 2 December 2027. AI embedded in regulated products moves to 2 August 2028. Both are now fixed calendar dates, not tied to the readiness of standards.
Not everything slips. Article 50 transparency duties (chatbot disclosure, AI-content labelling) still apply from August 2026, with until 2 December 2026 for machine-readable marking of systems already on the market. A new prohibition on non-consensual intimate imagery applies from 2 December 2026, and penalties and national enforcement start in August 2026.
The economic point: waiting is valuable when the date is uncertain. With fixed dates, that reason for waiting has largely gone. What remains uncertain is detail, not timing. Check the final published text with counsel.
Cardan-AI recommendation: use the delay to build the no-regret layer now: an inventory of AI systems, supplier clauses on documentation and logging, and an owner per system. Do not buy rule-specific certification yet.
Analysis by
Cardan-AI Intelligence
Our research and analysis unit, dedicated to applied AI for business, industry and regulatory compliance.
Let's talk about your next competitive edge
Thirty minutes to identify the two or three use cases in your operations that pay for themselves within the first year.