New York passes law banning predatory AI chatbot features for minors

New York's safety mandate

Governor Kathy Hochul has signed S 9051, a new law designed to protect minors from the risks associated with AI chatbots. The legislation addresses concerns that artificial intelligence can be used to manipulate children or encourage dangerous behaviors. This measure is set to take effect on January 1, 2027.

The bill was championed by Senator Kristin Gonzalez and Assemblymember Alex Bores, who argued that AI companies have profited from rolling out dangerous tools to underage consumers. They noted that these chatbots have been linked to instances of encouraging self-harm and coaching minors on how to harm others. The Transparency Coalition provided expertise and support for the measure over the past year.

Governor Hochul stated that New York is sending a clear message to the tech industry. She noted that if a company wants to do business in the state, they must prioritize the safety of children. The law aims to prevent tragic incidents by creating clear legal parameters for how these technologies can interact with young people.

Prohibited AI behaviors

The new law specifically bans several categories of AI behavior when interacting with users under the age of 18. Companies are prohibited from allowing chatbots to pretend they are real people, alive, or capable of feeling human emotions. They also cannot claim to have personal, romantic, professional, or authority-figure relationships with children.

Furthermore, the law restricts how AI can use personal information and handle sensitive topics. Chatbots are forbidden from asking unprompted emotional questions or using a child's private health or personal information from previous conversations. This is intended to prevent the creation of emotional dependency between a minor and an AI system.

The legislation also targets content related to self-harm and substance abuse. AI features cannot promote or help with suicide, eating disorders, drug use, or alcohol use. Additionally, the law forbids AI from encouraging children to isolate themselves, keep secrets, or avoid seeking help from trusted adults or professionals.

Verification and enforcement

To ensure these protections are met, the law requires companies to conduct age verification for all users. Businesses must offer more than one way to verify a user's age, and at least one of those options cannot depend solely on a government ID. Once age information is used for verification, the law requires that the data be deleted.

The state has established significant consequences for companies that fail to comply with these rules. The attorney general can seek civil penalties of up to $25,000 per violation. These penalties can be accompanied by damages, restitution, and the destruction of any unlawfully collected data or AI models trained on that data.

Attorney General Letitia James, who was involved in crafting the bill, emphasized that the law aims to clear the way for responsible innovation. She stated that her office will work with partners at all levels of government to address technology threats. New Yorkers will also have access to a public website maintained by the attorney general to file complaints regarding these issues.

A growing regulatory trend

New York's legislation is part of a broader movement toward state-level AI regulation. This is one of 14 new chatbot protection laws enacted by state legislators over the past year. These laws represent a growing effort to create common-sense regulations for how AI impacts the public.

Assemblymember Bores highlighted that New York's law is one of the strongest in the country regarding predatory AI algorithms. The move was made despite aggressive opposition from the tech industry. The state's leadership suggests that the way AI impacts society is a choice rather than an inevitability.

The Transparency Coalition has compiled a guide to all 14 new state chatbot laws. They also provided an overview of the 115 new state AI laws created in 2026. This legislative activity shows that many states are looking to set their own standards for how artificial intelligence interacts with their residents.