EU AI Act T-33 Days: Chatbot Disclosure and Deepfake Labels Become Law August 2
Summary: On August 2, 2026 — 33 days from today — the bulk of the EU AI Act takes effect, imposing disclosure, labeling, and audit requirements on anyone serving EU users.
Key Facts
- Article 50 (Transparency): Chatbot operators must inform users they are interacting with AI; AI-generated images, audio, and video must be identifiably labeled; deepfakes published to inform the public on matters of public interest require clear, visible marking
- High-Risk AI Systems: AI used in hiring, credit scoring, healthcare, law enforcement, migration, or critical infrastructure must complete conformity assessments, finalize technical documentation, affix CE marking, and register in the EU database — all by August 2
- Penalties for non-compliance: up to €35M or 7% of global annual revenue, whichever is higher — more punitive than GDPR
- A deadline extension for high-risk systems has been discussed but not enacted; enterprises should treat August 2 as the binding date
Why It Matters
Like GDPR, the EU AI Act applies wherever EU residents are served — meaning US and Korean companies are fully in scope. With 33 days left, any organization that hasn't started conformity work is running out of time. The transparency requirements alone affect virtually every deployed AI product.
Read More
- EU AI Act Transparency Obligations Guide — Sidley Austin
- U.S. Companies and the August 2026 Deadline — Holland & Knight