New York bill would ban AI chatbots from giving legal or medical advice
TL;DR:
- Senate Bill S7263 would prohibit AI chatbots from providing legal or medical advice in New York, applying the same licensing principles that prevent unlicensed individuals from practising law or medicine.
- Users could sue companies whose chatbots cross the line, and disclaimers saying “I’m not a doctor” would not shield providers from liability if the system still delivers prohibited advice.
- The bill passed unanimously out of the Senate’s Internet and Technology Committee, and sits alongside broader AI regulation efforts targeting child safety and synthetic media transparency.
New York lawmakers are preparing to draw a hard boundary around AI chatbot capabilities. Senate Bill S7263, which cleared the Senate’s Internet and Technology Committee with unanimous support, would ban AI chatbots from providing advice that substitutes for licensed legal or medical professionals.
The core principle is straightforward: if an individual cannot practise law or medicine without a licence, an AI system should not be able to either.
Disclaimers will not be enough
The bill addresses a loophole that many AI providers currently rely on. Chatbot providers would need to inform users they are interacting with AI rather than a human professional, but that disclosure alone would not provide legal protection. If a system delivers advice that effectively replaces a licensed professional’s guidance, the company behind it could still face liability.
Users would be able to file civil lawsuits and recover damages and legal fees if they prove a chatbot provided unauthorised professional advice. State Senator Kristen Gonzalez, who introduced the bill, framed it as a matter of accountability: “People deserve real care from real people.”
The grey area problem
Educational explanations about general concepts would remain permitted. The difficulty lies in distinguishing between a chatbot summarising publicly available medical information and one effectively diagnosing or advising on treatment. The same explanation could inform one user and influence another’s health decisions.
The bill forms part of a wider New York push on AI regulation, with companion legislation targeting child protection and generative AI transparency requirements.
For UK policymakers and AI developers, New York’s approach is worth watching. Technology regulation enacted in large US states tends to influence legislation elsewhere. The UK does not currently have equivalent restrictions on AI professional advice, but as healthcare AI tools from Amazon, OpenAI, and Anthropic expand, the question of where general information ends and professional advice begins will need answering on this side of the Atlantic too.