The EU has made its new AI transparency rules enforceable across the bloc as of August 2nd. Article 50 of the AI Act requires publishers to label AI‑generated content and inform users when they interact with a chatbot or agent instead of a real person. The Commission says these obligations “foster trust and integrity in the information ecosystem.”

The rules mandate an “AI” mark if machines assisted in creating authentic‑looking deepfakes—image, audio, video or text—and an “AI‑Generated” label for fully AI‑produced material, such as deep‑fake videos of politicians, AI‑composed music, or AI‑summar(end) news.

Providers of general‑purpose AI models must document training data, publish summaries and adopt copyright policies, with non‑compliance subject to enforcement by national authorities. Penalties reach €15 million or three percent of global revenue for companies and €750k for EU bodies; small firms receive proportional relief.

The EU’s move makes it the world’s leading AI regulator, sparking industry criticism that the rules may slow innovation. The Commission argues the harmonised framework simplifies development and protects health, safety, fundamental rights and democracy, making effective implementation a priority.