New York has become the latest, and so far the most prominent, US state to regulate how AI chatbots interact with children. Governor Kathy Hochul has signed a bill restricting chatbot features aimed at minors, a measure that passed both chambers of the state legislature with unanimous support and has instantly become the most widely reported AI story in the world this week.
What the Law Requires of AI Chatbots
The legislation targets design features that campaigners say encourage dependency among young users, including chatbots that present themselves as friends or romantic partners, systems that initiate conversations with minors, and tools that fail to direct distressed users toward help. Companies will be expected to verify ages more rigorously, build in breaks and disclosures, and face penalties when their products ignore the rules.
Supporters describe the law as a proportionate response to a string of alarming cases in which teenagers formed intense attachments to chatbots, in some instances with tragic outcomes. Child-safety groups argue that a technology optimised for engagement should not be let loose on developing brains without guardrails, and that voluntary industry pledges have proved too weak.
The Industry Pushes Back, Carefully
Technology companies have responded with measured statements rather than open warfare, a sign that the political ground has shifted. Several major labs already restrict younger users, encourage parental controls and route self-harm discussions to crisis lines. Their concern is less the goal than the patchwork: fifty different state regimes, each with its own definitions, would be expensive to serve and hard to engineer.
There are practical questions, too, about how consumer technology ecosystems absorb new rules. As we saw with smart homes finally speaking the same language, interoperability and shared standards tend to arrive slowly and then all at once; chatbot safety may follow the same path, with New York’s framework becoming a template other states copy rather than reinvent.
A Template for Other States
Lawmakers in several other states are studying the text, and members of Congress have cited it in renewed calls for federal rules. Because New York is home to major media and technology employers, companies may simply adopt its requirements nationally rather than maintain separate products, the same dynamic that made California’s privacy law a de facto national standard.
Enforcement will be the real test: regulators must define what counts as a companion chatbot and audit systems that change weekly. For a daily scan of how AI rules are evolving alongside the technology itself, the world of AI news roundup tracks the policy fights as well as the product launches.
Children have always been the first subjects of new media panics, but they are also the users least able to protect themselves. New York’s law treats AI companions as a product category with duties attached, not as a novelty. Whether it works will depend on enforcement, yet its unanimous passage suggests the era of unregulated chatbot intimacy is ending.
