The EU AI Omnibus entered into force yesterday (on 27 July 2026), introducing targeted simplifications and extended compliance timelines under the EU AI Act.

The changes aim to reduce administrative burdens and provide greater legal certainty for companies developing or deploying AI systems in the EU, while maintaining the existing safeguards on safety and fundamental rights.

Key changes include:

• Extended timelines for high-risk AI systems:
  • Annex III high-risk AI systems: requirements will apply from 2 December 2027.
  • High-risk AI systems embedded in regulated physical products under Annex I: requirements will apply from 2 August 2028.
• Broader support for growing companies: Certain simplified requirements and support measures previously limited to SMEs will also apply to small mid-cap companies.
• More testing opportunities: Access to regulatory sandboxes will be expanded, including through a new EU-level AI regulatory sandbox.
• Reduced administrative obligations: The Omnibus simplifies certain requirements relating to AI literacy, registration in the EU AI database and conformity assessment procedures.
• Clarified governance and enforcement: The powers of the EU AI Office are extended for certain AI systems, including systems based on general-purpose AI models.

As discussed, the extended deadline for AI systems embedded in physical products is particularly relevant to companies developing connected or AI-enabled equipment.

More broadly, companies using AI for recruitment, workforce management, consumer services, product development, marketing or other business processes should continue assessing whether their systems fall within the AI Act’s risk categories.