On May 7, 2026, the European Parliament and the Council of the European Union reached a provisional agreement on AI Omnibus (Omnibus VII), aiming to simplify and optimize parts of the EU AI Act implementation arrangements. For manufacturers, this does not mean the AI Act is being fully postponed, but rather that compliance obligations for some high-risk AI systems are being reclassified and implemented in phases, with major changes including:

• Postponed from August 2, 2026 to December 2, 2026: Transparency obligations such as labeling and watermarks for AI-generated content are expected to apply; Newly prohibits AI-generated child sexual abuse material or involuntary intimacy/sexually explicit content that identifies individuals;
• Postponed from August 2, 2026 to December 2, 2027: Some Annex III high-risk AI system obligations are expected to apply, such as biometrics, critical infrastructure, education, employment, law enforcement, border management, and other scenarios;
• Postponed from August 2, 2027 to August 2, 2028: High-risk AI systems embedded in products and used as security components, especially those regulated by EU industry safety regulations, are expected to be subject to relevant obligations.

The above changes were approved by the European Parliament on June 16, 2026, and after adoption by the Council of the European Union, they are expected to be officially published to the OJ by August 2, 2026.

Click this link to view the European Parliament's press release on the approval of the revision of the AI Act.