European Sovereignty Monitor

Policy and law

EU AI Act bans on unacceptable risk AI practices, literacy duties start

The EU AI Act's ban on eight unacceptable risk AI practices, including social scoring, untargeted biometric scraping and workplace emotion recognition, took effect on 2 February 2025, alongside new AI literacy duties for providers and deployers.

Reported AI-drafted European Commission, AI Act regulatory framework guidance (digital-strategy.ec.europa.eu)

What it does not test. It does not show how national market surveillance authorities are enforcing the ban in practice.

How this item counts. This item is dated before the index starts on 1 January 2026, so it does not count in the European Sovereignty Index.

How this item was made. Drafted by an AI system (Claude (Anthropic)) from the source cited above and published under the editor’s responsibility without a person reading it first. How the record is made

What the tags mean: VerifiedReportedAI-draftedAI-drafted, editor-readHuman-writtenWhat the tags mean

Previously in Policy and law

VerifiedOpinionAI-drafted DIGITALEUROPE's Bonefeld-Dahl: sovereignty rules need high fences only for sensitive fields ReportedOpinionAbroadAI-drafted FFO says Ofcom cannot enforce Online Safety Act fines on US websites VerifiedAI-drafted EU-backed OSOR report finds national open source policy now a sovereignty tool

All Policy and law news →