EU AI Act deadline: does August 2, 2026 still apply?
Why there are two dates
The AI Act as adopted in 2024 set August 2, 2026 as the date most high-risk obligations begin: conformity assessments, technical documentation, CE marking, registration in the EU database, human oversight, and the rest of Articles 8 through 15. In November 2025 the European Commission proposed the Digital Omnibus on AI, a simplification package that, among other changes, defers the high-risk compliance date to December 2, 2027.
That proposal has now cleared both legislators: Parliament endorsed it on June 16, 2026, and the Council gave final approval on June 29, 2026. What remains is publication in the Official Journal, after which the amending act enters into force. This is normally a formality measured in weeks — but it has not happened yet, and EU law is what is published, not what is agreed.
What this means in practice
If you deploy or provide AI in hiring, credit scoring, insurance pricing, education assessment, critical infrastructure, biometrics, or the other Annex III areas, your legal deadline today is still August 2, 2026. The near-certain deferral gives you breathing room in expectation, not in law.
The prudent posture: treat the deferral as schedule relief for the expensive tail of compliance (notified-body conformity assessment, CE marking) while continuing the work that is valuable under either date — knowing your risk classification, building the technical documentation skeleton, and assigning human oversight. Companies that dismantled their AI Act programs on the news of the omnibus vote are exposed if publication slips past August 2.
Deadlines in this space have moved twice in twelve months (the omnibus also touched CSRD and the EUDR saw its own postponement). Any compliance answer that doesn't carry a verification date and a source is stale on arrival.
What is already in force and NOT deferred
Prohibited practices (Article 5) have been banned since February 2, 2025: social scoring, manipulative techniques causing significant harm, untargeted facial-image scraping, emotion recognition in workplaces and schools (outside safety uses), and real-time remote biometric identification in public spaces outside narrow law-enforcement carve-outs. Penalties reach €35M or 7% of worldwide turnover.
General-purpose AI model obligations took effect August 2, 2025: transparency documentation, a copyright policy, and training-data summaries for model providers, with a systemic-risk tier carrying more. None of this is touched by the high-risk deferral.
The AI literacy obligation (Article 4) also already applies — staff working with AI systems need appropriate training, a cheap requirement that regulators can check first.
How to check your own exposure
Three questions determine most outcomes. First: does your AI touch people in the EU — users, candidates, customers, data subjects? The Act applies by effect, not by where your company sits. Second: does the use case fall in Annex III (employment, credit, insurance, education, essential services, biometrics, critical infrastructure, law enforcement, migration, justice)? Third: what is your role — provider obligations are far heavier than deployer obligations, and importers and distributors carry verification duties.
Our /api/comply/ai-act endpoint runs this classification deterministically — Annex III and Article 5 surfaces, role obligations with article citations, and the current deadline state tracked to Official Journal status — for $0.25 a call. The /api/comply/deadlines endpoint ($0.05) tracks the publication status itself, so an agent can re-check weekly for pennies.
GET https://compliancepulse.theaslangroupllc.com/api/comply/ai-act — x402 pay-per-query, no API key. See llms.txt.FAQ
Has the EU AI Act high-risk deadline been officially delayed?
Approved but not yet legally effective: Parliament (June 16, 2026) and Council (June 29, 2026) approved deferring high-risk obligations to December 2, 2027, but until the amending act is published in the Official Journal, August 2, 2026 remains the binding date.
Are any AI Act rules already in force?
Yes. Prohibited practices since February 2, 2025 (penalties to €35M/7%), GPAI model obligations since August 2, 2025, and AI literacy duties. None of these are affected by the high-risk deferral.
Does the AI Act apply to non-EU companies?
Yes — it applies to providers and deployers wherever established, when the AI system is placed on the EU market or its output affects people in the EU.
What are the penalties for high-risk non-compliance?
Up to €15M or 3% of worldwide annual turnover for most violations; prohibited practices reach €35M or 7%.