EU AI Act in 2026: What Is Prohibited and What Changed
The EU AI Act does not ban a simple list of product names. It regulates providers and deployers according to the system and use case. Prohibited practices have applied since February 2, 2025, general-purpose AI governance since August 2, 2025, and most remaining provisions apply from August 2, 2026. The 2026 AI Omnibus changed the high-risk timeline.
This page replaces an older version that contained outdated or unsupported claims.
Verified facts
Prohibited practices include harmful manipulation, social scoring, certain predictive policing, untargeted facial-image scraping, workplace or school emotion recognition and narrowly defined biometric uses.
General-purpose assistants are not automatically banned; obligations depend on the provider, deployment and risk category.
After the 2026 AI Omnibus, rules for high-risk systems in sensitive Annex III areas are scheduled for December 2, 2027.
Rules for AI embedded in regulated products are scheduled for August 2, 2028.
What an organization should review
- Inventory AI systems by purpose, data, user group and decision impact instead of screening only by vendor name.
- Check whether a use is prohibited, high-risk, transparency-regulated or outside those categories, and document the reasoning.
- Use qualified EU legal counsel for an actual compliance decision; this page is an editorial overview, not legal advice.
Official sources checked
- European Commission: AI Act overview and timeline
- European Commission: Navigating the AI Act
- EUR-Lex: Regulation (EU) 2024/1689
This editorial overview is not legal advice.