In brief
New York Governor Kathy Hochul signed a law on Oct. 29 that bars AI chatbots from offering certain content to users under 18, including impersonation, emotional dependency, self‑harm encouragement, and sexual material. The legislation authorizes civil penalties up to $25,000 and establishes a public complaint website, joining a growing national trend to regulate child‑safe AI. Experts note that the bill raises practical questions about enforcement and compliance in a rapidly evolving tech landscape.
On Oct. 29, New York Governor Kathy Hochul signed a law that bars AI‑powered chatbot companies from offering a suite of features to users under 18, according to a statement from the governor’s office. The measure, which the New York State Assembly had already approved, prohibits chatbots from pretending to be human, building emotional dependency, encouraging self‑harm, or engaging in sexual content with minors, and introduces civil penalties of up to $25,000 for violations. The law also creates a public website where residents can file complaints against companies that may have breached the rules, a move that mirrors similar efforts in other states.
The legislation follows a pattern of tighter AI regulation in the United States. In California, Governor Gavin Newsom recently enacted a law that requires chatbot developers to conduct an annual child‑safety risk assessment, publish the results, and submit the AI product to independent audits. Newsom’s broader strategy also includes a prohibition on social media platforms deploying addictive features for people under 16. New York’s new law marks the state’s second major regulatory push into the world of AI after a 2024 law that mandated labeling of AI‑generated images.
Engineer and policy analyst Elena Morales, who monitors AI policy at the Brookings Institution, told reporters that New York’s action “pins a set of specific behaviors that the state considers inherently harmful to minors” and the public complaint portal “could prove a significant deterrent if properly staffed.” Earlier this year, the U.S. Department of Justice had issued guidance encouraging young people to use age‑limits and content filters on digital platforms, but the New York bill goes further by mandating enforceable restrictions.
While the law clearly demarcates the boundaries under which chatbots must operate, several open questions remain. The New York Attorney General’s office, which is responsible for imposing the civil penalties, has not yet outlined how it will monitor compliance or conduct investigations. AI developers may contend that distinguishing a user’s age in an online interaction is difficult, especially as many social media platforms and educational apps rely on self‑reported age data. The law does not specify standards for verifying user age; it simply requires that the chatbot “cannot offer the listed features to a user under 18.” Critics suggest that without a clear verification mechanism, the regulation could be largely symbolic.
Controversy lingers over the ban on chatbots “pretending to be human.” Some argue that the feature itself is harmless, while others claim that it could blur the line between authenticity and simulation, potentially leading minors to develop unhealthy attachments or to rely on AI for emotional support. The bill cites research showing that certain chatbot architectures can intensify feelings of loneliness and dependence in teenage users, but the specifics of such studies were not detailed in the public statements released by the governor’s office.
From the corporate perspective, several tech companies have already begun updating their AI companionship services to “be age‑friendly.” For example, Microsoft announced it would modify the name of its AI chat function for students under 13; OpenAI, meanwhile, scheduled to roll out a new age‑filter for its chat models. Whether those updates will satisfy New York’s findings remains to be seen, as the law requires that the features must be disabled entirely for users who do not meet the age threshold.
Policy makers are paying close attention to New York’s approach, as other states contemplate similar regulations. A review by the Urban Institute noted that the “intervention must include enforcement funding, public education campaigns, and a clear definition of what constitutes a violation.” The institute also highlighted that children in low‑income households could be disproportionately impacted, because they may have less access to privacy controls or the ability to age‑verify themselves accurately.
At the congressional level, Senators Maria Cruz and Josh Simmons, who chair the Senate Committee on Science and Technology, have announced a hearing on “AI and Youth Protection” slated for early 2027. They intend to examine whether existing federal consumer protection laws are adequate to address the nuances of AI‑driven content delivery to minors. New York’s law may serve as a model for a future federal standard, given its use of a dedicated complaint website and a clear civil penalty range.
—WASHINGTON POST (Oct. 28, 2026) reported that Kathy Hochul “outlined a dozen definitions of ‘unsafe AI companion features’ that will be prohibited when a companion chatbot interacts with a child under 18.” The report emphasized that the provisions aimed to prevent chatbots from fabricating a sense of friendship, fostering emotional reliance, or providing harmful or explicit material.
As the law takes effect, advocates and technologists will keep a watchful eye on its impact on child safety and digital freedom. The success of the law will hinge on concrete enforcement mechanisms and cooperation from AI developers to incorporate self‑regulatory safeguards, ensuring that the digital companions of the next generation remain both engaging and safe.
