NYC Proposes First-in-Nation Rule Forcing AI Chatbots to Hand Off to Humans
New York City's Department of Consumer and Worker Protection on October 20 proposed what it is calling the first rule of its kind in the country: a regulation that would require any business operating in the city to give customers a fast, reliable way to reach a human being whenever an automated chatbot — AI-driven or otherwise — fails to resolve their problem.
The rule targets what DCWP Commissioner Samuel Levine described as a customer-service 'doom loop,' in which companies route complaints through bots that repeat unhelpful menus, surface irrelevant or incorrect answers, or quietly discourage refunds and complaints. Under the proposal, businesses would have to disclose when a customer is interacting with an automated system, offer a one-touch option to escalate to a person, and respond in a 'reasonable and timely' manner. Companies that fail to comply could not shift blame onto the underlying technology.
Penalties scale with repeat violations
Civil penalties start at $525 for a first violation, rise to $1,050 for a second, and reach $3,500 for a third or subsequent offense — a structure explicitly designed to punish companies that treat chatbot-caused frustration as a cost of doing business rather than a compliance risk. Consumers harmed by a violation would also be able to seek restitution. The rule would cover any company doing business in New York City regardless of where it is headquartered, though it would not reach sectors under exclusive federal jurisdiction, such as airlines and national banks, since DCWP is relying on the city's existing deceptive-trade-practices authority rather than new state or federal legislation.
A 30-day public comment period is now open, with a hearing scheduled for November 19. How quickly the rule takes effect afterward depends on comment volume, and the proposal includes a transition window for companies to retool their systems.
Part of a wider, uncoordinated patchwork
The move lands amid a scramble of state-level activity that has outpaced Washington: lawmakers in 37 states have introduced more than 120 chatbot-related bills this year, according to the Future of Privacy Forum, and 18 states now have chatbot-specific laws on the books covering disclosure, crisis protocols, or protections for minors. California's Right to Human Customer Service Act, which takes effect January 1, 2027, applies only to companies with more than $500 million in national annual revenue and requires a good-faith effort to connect a caller to a person within 15 minutes or an appointment within one business day. New York City's version is broader in reach — it applies to any business operating in the city regardless of size — but narrower in mechanism, since it works through consumer-protection fines rather than a statutory right to a human agent.
Mayor Zohran Mamdani's administration is framing the rule explicitly as a response to federal inaction, arguing Washington has not moved fast enough to protect consumers from AI-driven customer service failures. Levine has said he hopes other jurisdictions adopt similar rules, which would effectively export New York's enforcement model to other cities without requiring any new state or federal statute. That approach — using existing consumer-protection law rather than AI-specific legislation — offers a template other municipalities could copy quickly, in contrast to the slower, often-stalled path of state legislatures and Congress on broader AI regulation.
The timing also reflects shifting public sentiment: Pew Research Center polling cited alongside the proposal found that roughly half of U.S. adults now use AI chatbots, up from about a third in 2024, even as Americans remain broadly skeptical of AI's overall impact — a gap between adoption and trust that city regulators are now moving to address at the point of greatest consumer friction: the customer-service call that never reaches a person.