New York has become the latest — and one of the largest — jurisdictions to regulate artificial intelligence companions, after Governor Kathy Hochul signed a bill restricting how AI chatbots can interact with minors. The measure, which passed both chambers of the state legislature with unanimous support, was the most widely carried AI story of the week. Our technology coverage examines what the law requires and how the industry is responding.
What the New Law Requires
The legislation places obligations on companies whose chatbots are used by children, targeting the design features that encourage prolonged, emotionally dependent relationships between minors and AI systems. Supporters describe it as a child-safety measure modelled on earlier social media rules: platforms must build in safeguards rather than leaving parents to police every interaction alone.
The unanimous votes in both chambers underline how completely the politics of child online safety have shifted. Measures that once split legislatures along party lines now pass without dissent, as lawmakers respond to parents’ concerns about screen time, mental health and the persuasive design of AI companions.
Why AI Chatbots Became a Political Target
AI chatbots differ from earlier social platforms in ways that worry researchers: they talk back, they remember, and they can simulate friendship or even romance. High-profile cases involving teenagers and companion apps have pushed the issue up the agenda, and regulators increasingly argue that a conversational agent needs different rules from a static website.
As tracked by daily AI news roundups, the New York law was reported by more than forty outlets, making it the week’s defining AI policy story. Other states are expected to study its text closely, and similar proposals are already circulating in several legislatures.
The Industry’s Dilemma
For AI companies, state-by-state child-safety laws create a patchwork of compliance duties on top of fast-moving product development. Age verification, parental controls and limits on engagement features are technically feasible, executives concede, but designing them without locking out legitimate users — or collecting even more data about children — is a genuine engineering and privacy trade-off.
Some firms are likely to apply the New York standard nationally rather than maintain different products per state, which would make the law a de facto national rule. Others may challenge aspects of it in court. Either way, the era of launching AI companions with minimal guardrails is ending.
New York’s chatbot law is unlikely to be the last word on children and AI, but it may be the template. As the technology grows more personable, the rules governing who it can talk to — and how — are becoming a central battleground of tech policy.
