EU begins enforcing AI Act transparency rules

The European Commission's AI Office and national authorities began enforcing new transparency requirements under the EU's Artificial Intelligence Act. Chatbots and other interactive AI systems must now tell users they are dealing with AI rather than a human, a rule covering voice assistants, AI companions, social media bots and agentic systems that contact people autonomously. AI-generated or altered content, including deepfake images, video and audio, must carry machine-readable marks identifying it as synthetic.

The obligations apply to both providers that build AI systems and deployers that use them, from a bank running a third-party chatbot to a social platform hosting synthetic media. Non-compliance can draw fines of up to €15 million or 3 percent of global annual turnover, whichever is higher. Tougher rules for high-risk AI systems remain deferred, to August 2028 for those embedded in regulated products and December 2027 for sensitive uses such as recruitment and credit scoring.

The rules ask AI systems to disclose what they are, not govern what they do, so most of the near-term compliance burden falls on interface design rather than model behaviour.