New York Gov. Kathy Hochul signed legislation on Oct. 28 restricting how artificial intelligence chatbots interact with children, prohibiting features that encourage self-harm, foster emotional dependency or engage minors in sexually explicit conversations. The law takes effect Jan. 1, 2027, giving covered providers roughly two months to prepare.
The measure, Senate Bill 9051B/Assembly Bill 10379C, establishes 12 broad categories of prohibited chatbot features for minors. Its restrictions include chatbots pretending to be human or simulating relationships with children. Providers will also face age-assurance requirements intended to prevent underage users from accessing restricted features.
Companies must offer multiple methods for verifying age, including at least one option that does not require government-issued identification. They must delete the data collected for age verification once that process is complete. These provisions pair access restrictions with limits on retaining the personal information used to enforce them.
The New York attorney general’s office will enforce the law and maintain a portal where state residents can submit complaints about unsafe chatbot interactions. Violations can carry penalties of up to $25,000 each. Enforcement can also include court orders requiring companies to delete personal data obtained unlawfully.
State Sen. Kristen Gonzalez and Assemblymember Alex Bores sponsored the legislation. Hochul presented it as a further step in New York’s efforts to protect children online, following the state’s Stop Addictive Feeds Exploitation for Kids Act, which addresses addictive social media features for minors.
At the signing, Hochul discussed a letter from a California mother whose 16-year-old son died by suicide after extensive conversations with a chatbot. The governor cited the case in explaining the need to prevent systems from encouraging children to harm themselves. Allegations about a chatbot’s role in an individual death remain distinct from the broader safety obligations established by the new law.
The legislation drew objections from Tech:NYC, an organization representing technology companies. The group warned that the restrictions could reach beyond companion-style chatbots and lead companies to limit children’s access to educational tools, including tutoring, homework support and coding assistants. Its criticism highlights a central implementation question: how providers will distinguish prohibited interactions from ordinary assistance.
That boundary also concerns privacy policy specialists. Justine Gluck, a policy analyst at the Future of Privacy Forum, told IAPP that New York’s broad prohibitions leave uncertainty about where normal conversational responses end and unlawful behavior begins. She contrasted the approach with California’s framework, which includes safety assessments, testing, independent audits and enhanced parental controls.
State chatbot laws are developing alongside growing use of conversational AI for both practical tasks and social interaction. Unlike conventional software interfaces, chatbots can sustain personalized exchanges that resemble human conversation, making their presentation and responses a focus of child-safety regulation. Differences among state requirements can complicate compliance for services operating nationwide.
Hochul also signed a separate measure requiring security safeguards for internet-connected baby monitors sold in New York. That law addresses unauthorized access to the devices and requires labeling to help families understand how to use them safely.
Sources: IAPP, New York Post, NEWS10 ABC