EU AI Act Amendments Push High-Risk Deadlines, Add Prohibitions
Any company developing, deploying, or integrating AI in the EU must reassess compliance timelines and prohibited-use policies following the July 2026 AI Act amendments.
The EU AI Act Amendment, effective 27 July 2026, extends compliance deadlines for high-risk AI systems: Article 6(2)/Annex III systems now have until 2 December 2027, while Article 6(1)/Annex I systems tied to product safety legislation have until 2 August 2028. The Commission cited delayed harmonized standards and national competent authority readiness as justification. Transparency obligations for synthetic audio, image, video, and text generators (including GPAI) are pushed to 2 December 2026; other transparency rules remain on the original 2 August 2026 date. The prohibited-AI list expands to cover systems generating non-consensual intimate imagery, effective 2 December 2026. AI literacy duties softened from 'ensure' to 'support' development. New supply-chain cooperation duties and penalties apply when original providers fail to assist downstream providers. In-house counsel should update compliance roadmaps, audit high-risk classifications, and review synthetic-content and intimate-imagery use cases.