The simplest AI rule is also spreading fastest: tell people when they are talking to a machine. For chatbots that act as companions, lawmakers are adding safety duties too.
European Union
Under Article 50 of the AI Act, which has applied since 2 August 2026, providers must design AI systems that interact with people so that people are told they are dealing with AI, unless it is obvious. Deepfakes and AI-generated text published on matters of public interest must be labelled.
United States
There is no federal chatbot law. The states are moving:
New York has required operators of AI companions to have protocols for suicidal ideation and self-harm, and to remind users they are not human, since November 2025.
California’s companion chatbot law (SB 243) has applied since 1 January 2026: disclosure that the chatbot is not human, crisis protocols, and extra protections for minors.
Tennessee bars AI from presenting itself as a licensed mental health professional, from July 2026.
Oregon and Washington companion chatbot laws start on 1 January 2027, and Nebraska and Idaho conversational AI laws on 1 July 2027. Oregon’s reportedly includes a private right of action with statutory damages.
United Kingdom
There is no chatbot-specific law. Consumer protection law prohibits misleading practices, and Ofcom has said the Online Safety Act can apply to generative AI chatbots that let users share content or that search the web.
A one-line disclosure is cheap. A missing one is now a breach in several places at once.
What to do
Add a clear AI disclosure to every customer-facing chatbot. If your chatbot could be used for companionship or by children, check the state laws on crisis handling and minors before launch.
This article is general information on the law as at November 2026. It is not legal advice and has not been updated since that date. See the terms of use.