Antitrust Law Daily Wrap Up, PRIVACY NEWS: Bills introduced in Senate, House to protect kids from chatbots, (Jul 31, 2026)

By Lynn Stanton, TR Daily

The bills follow on the heels of several state laws seeking to limit AI chatbot use by children.

Bills have been introduced in both the Senate and the House with the aim of protecting children from the potential mental health harms and privacy threats of AI (artificial intelligence) chatbots.

The bipartisan Children Harmed by AI Technology (CHAT) Act 2.0, introduced by Sens. Jon Husted (R., Ohio) and Andy Kim (D., N.J.), is aimed at companion chatbots with educational purposes or that simulate interpersonal or emotional interactions over multiple interactions. It would require the owner or operators of a covered chatbot to provide age assurance that includes the age and birthdate of the user.

CHAT Act 2.0 would prohibit owners or operators of covered chatbots from engaging in child-targeted advertising or the sale or sharing of the data of minor users without parental consent. It would also impose requirements for documenting the data used to train the chatbot and how it collects data from minor users. It would require chatbots to comply with health privacy safeguards to be established by the Federal Trade Commission.

It would require the owners or operators of covered chatbots to provide clear instructions on how to delete the history of conversations with minor users.

CHAT Act 2.0 would require covered chatbot owners and operators to prevent companion chatbots from encouraging or normalizing suicide, self-harm, or physical violence; engaging in sexually explicit communication with a minor user; creating emotional dependence; simulating a romantic relationship with a minor user; “engaging in conduct likely to cause a minor to believe such chatbot is an emotional partner; or “facilitating the promotion, creation, solicitation, or distribution of child sexual abuse material.”

It also would require clear disclosures that the chatbot is not human. Educational chatbots that cannot provide companionship or therapeutic relationships would have to make such disclosures at least every three hours during a continuing session with a minor user, while more frequent notifications and “nudge-out” notifications would be required for chatbots that offer companionship or therapeutic relationships. Companionship and therapeutic chatbots would have to make referrals to the 988 Suicide & Crisis Lifeline and other crisis service providers, as appropriate, and to notify parents of suicidal ideation.

The bill would direct the National Institute of Standards and Technology to conduct a study of the impacts of companionship and therapeutic chatbots on self-harm and suicide; manipulation and emotional dependency; and exploitation of minor users.

The bill would authorize enforcement by the FTC of violations of CHAT Act 2.0 as unfair or deceptive acts or practices. State attorneys general could bring civil actions under the legislation but would have to inform the FTC of their actions.

In the House, Reps. Becca Balint (D., Vt.), Chris Deluzio (D., Pa.), Tom Suozzi (D., N.Y.), and Bonnie Watson Coleman (D., N.J.) have introduced the Addictive Design Act, which would require “companies that use age verification to minimize the data they collect and delete that data within 24 hours,” according to a press release.

It would also “[b]an AI chatbot companies from letting youth access an AI chatbot with design features known to increase addiction, unhealthy emotional attachment, and compulsive use,” the press release said.

It would “[e]stablish an interagency task force to make recommendations for addressing the mental health risks associated with AI chatbot use in youth, including unhealthy emotional attachment and addiction”; authorize $3 million to fund research into mental health risks posed by minors’ use of AI chatbots; authorize $5 million to fund outreach to parents and educators to help them identify mental health warning signs related to minors’ use of AI chatbots; and require covered operators of chatbots to provide data to grant-funded researchers who are studying the mental health effects of AI chatbot use.

It would “[a]uthorize civil penalties of up to $5,000 per youth impacted for companies that make AI chatbots with addictive design features available to minors.”

News: AINews Privacy