New York has enacted a law restricting the behavior of AI companion chatbots when they interact with minors, setting a new standard for how states seek to regulate conversational artificial intelligence. Governor Kathy Hochul signed the measure on October 28, and its requirements are scheduled to take effect on January 1, 2027.
Which chatbot behaviors are restricted?
The legislation targets features considered unsafe for people younger than 18. It prohibits chatbot interactions that present a machine as a real person with emotions or a personal relationship, promote secrecy from trusted adults or encourage harmful behavior. The law also restricts certain sexually explicit or manipulative exchanges.
Its focus is on chatbot design and the nature of the interaction, rather than simply requiring developers to add a warning that users are speaking with artificial intelligence. The state Senate’s legislative record confirms that the measure was signed into law.
Potential penalties for developers
The New York attorney general may pursue enforcement against operators whose products violate the rules. Potential consequences include financial penalties and orders affecting the operation of an unsafe product. Exactly how individual cases are handled will depend on implementation and legal interpretation.
Developers will need to examine whether their services qualify as AI companions and how the rules apply when they know a user is underage. Age-assurance systems and content safeguards are likely to form part of compliance planning, although the statute’s scope is more specific than a blanket ban on minors using all AI tools.
Part of a broader state-level shift
New York follows other states, including California, in seeking stronger protections for young chatbot users. Advocacy organizations have supported the move, arguing that systems designed to maintain prolonged personal conversations can create particular risks for children.
Technology policy analysts have also cautioned that broad language could create uncertainty around routine conversations and services that are not intended to function as companions. Clear guidance on definitions and enforcement will therefore matter.
What companies must watch
The new law gives AI developers a limited period to evaluate affected features before the January effective date. Companies operating nationwide face an additional challenge: state laws can differ in their definitions of unsafe conduct and compliance requirements.
The measure is an important test of whether lawmakers can regulate specific chatbot behavior while leaving space for useful educational, creative and informational AI services.
Sources: New York State Legislature, signed October 28; Associated Press, October 29; Future of Privacy Forum, October 29.
