CSOAI - the 5-minute version
The EU AI Act, explained
The first comprehensive AI law, in plain English. It sorts AI into four risk tiers and regulates each differently. Here is the whole thing in five minutes.
You are interacting with an AI system.
The embedded 'Ask the Sovereign' panel (SovereignSpot) sends questions to the live Sovereign chat endpoint (os.meok.ai/api/chat), where a model writes the answer. The Art 50(1) notice for this surface is registered here and being wired; until the component ships, this registry entry is the disclosure.
Disclosed under EU AI Act Article 50(1). Every surface and its classification
The four risk tiers
Banned outright - social scoring, manipulative AI, untargeted facial scraping. In force since Feb 2025.
Heavily regulated - biometrics, hiring, credit, healthcare, justice. Risk management, oversight, conformity. Phasing to 2027.
Transparency duties - tell people they are dealing with AI; label synthetic media. From 2 Aug 2026.
Most AI - spam filters, recommendation engines. No mandatory obligations; voluntary codes encouraged.
Providers + deployers touching the EU market - including non-EU companies.
Bans Feb 2025; GPAI Aug 2025; transparency 2 Aug 2026; high-risk through 2027.
Up to EUR 35m or 7% of global turnover.