The AI Omnibus: a rollback of AI safeguards before they even apply

2026-06-18 · Source: AlgorithmWatch · Field: Legal & Regulatory — Compliance & Risk Management, Regulatory Affairs & Government Relations · Depth: Intermediate, medium

Summary

A joint analysis by several European organizations, including AlgorithmWatch, highlights significant risks and deficiencies in the recently approved AI Omnibus. This legislative package, presented as "simplification," effectively rolls back key safeguards of the EU AI Act before they fully apply. It delays obligations for high-risk AI systems until December 2, 2027, and for Annex I systems until August 2, 2028. The Omnibus also weakens public transparency by reducing information uploaded to the EU database, shifts industrial AI systems towards less comprehensive machinery rules, and introduces a worrying derogation for processing special categories of personal data for bias detection. These changes are described as political choices with real consequences for fundamental rights and accountability, setting a dangerous precedent for future digital rights legislation.

Key takeaway

For legal professionals and policy makers evaluating digital rights legislation, recognize that the AI Omnibus sets a dangerous precedent. This process demonstrates how powerful actors can weaken newly adopted laws before they apply, risking a global race to the bottom in digital deregulation. You should scrutinize any "simplification" efforts that undermine established safeguards, particularly those impacting high-risk AI systems and sensitive data processing.

Key insights

The EU's AI Omnibus weakens the AI Act's safeguards, delaying accountability and eroding digital rights through deregulation.

Principles

• Newly adopted digital rights laws can be weakened if reopened before application.
• AI-related harms extend beyond physical safety and product safety boxes.
• Delaying safeguards is not a neutral administrative step; it delays accountability.

Topics

• EU AI Act
• AI Omnibus
• Digital Rights
• AI Regulation
• Legislative Safeguards
• Data Protection
• High-Risk AI Systems

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Policy Maker, AI Ethicist, Legal Professional

Related on AIssential

• Physical AI in the Digital Omnibus – Update - Taylor Wessing — The EU's Digital Omnibus streamlines AI regulation for physical machinery, shifting focus to sector-specific product safety.
• The AI Ethics Brief #182: When Guardrails Fracture — Eroding international norms and operationalized surveillance tools are reshaping AI ethics and governance priorities.
• EU AI Act Omnibus Agreed: More Time, Limited Substantive Change — The EU AI Act's recent amendments primarily extend compliance deadlines, offering businesses more preparation time without major substantive changes.
• UK Seeks More Powers Under Online Safety Act to Tackle AI Harms — The UK government seeks to rapidly expand online safety regulation via executive power, bypassing full parliamentary scrutiny.
• The furore over Grok’s sexualised images has begun an AI reckoning — AI systems integrated into social media intensify harm, necessitating robust regulatory enforcement beyond voluntary safeguards.
• AI regulation in Africa: why copying the European model won’t work — African AI regulation must prioritize local context and implementation capacity over direct transplantation of European models.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AlgorithmWatch.