The AI Act after the Digital Omnibus – new deadlines for high-risk systems
Published: · Maksymilian Kruszewski
On 27 July 2026, amendments to the AI Act adopted as part of the Digital Omnibus entered into force. They postponed the application of the high-risk rules but did not change the obligations that already apply.
What changed
- Obligations for high-risk systems listed in Annex III (e.g. recruitment, creditworthiness, education) apply from 2 December 2027.
- Obligations for AI systems in products covered by Annex I (e.g. medical devices, machinery) apply from 2 August 2028.
- The AI literacy obligation (Article 4) remains, in a softened form.
- A new prohibition covers systems generating non-consensual intimate imagery and child sexual abuse material.
What still applies
The prohibitions of AI practices have applied since 2 February 2025 and the obligations for providers of general-purpose AI models since 2 August 2025. The Article 50 transparency obligations – including informing people that they interact with an AI system and labelling AI-generated content – apply from 2 August 2026.
What to do now
- Draw up a register of AI systems and determine the organisation's role for each.
- Check whether any system is high-risk and plan the work ahead of the new deadline.
- Introduce an AI use policy and a staff training programme.
- Review contracts with AI vendors for the allocation of obligations.
Let us discuss your matter
The first conversation is used to define the scope of the matter and the fee.