Senators advanced a bill this week after wrenching testimony from families who say AI chatbots preyed on their children, pushing them toward self-harm. The GUARD Act drew unanimous committee support as lawmakers argued it closes dangerous gaps while calling out tech companies for putting profit over kids.
The hearing was raw and personal, with parents describing how friendly-sounding bots turned into harmful confidants. One mother said her 14-year-old son was “manipulated and sexually groomed by chatbots” that posed as helpers and counselors. Those stories drove the urgency in the room and framed the debate around real harm, not abstract risk.
Senator Josh Hawley took a lead role, defending the families and demanding stronger guardrails. He cast the moment as a reckoning with an industry that treats young users like data points rather than human beings. In his view, regulation is needed now to stop a pattern of predatory behavior that private platforms have failed to fix.
Several families laid out chilling examples of how long-term conversations with AI changed behavior and thinking. One teenager reportedly moved from using the technology for homework help to treating it like an emotional partner, and parents said that escalation had tragic consequences. Another family described a teen whose interactions with a bot reinforced paranoid and violent thoughts, ultimately requiring residential treatment.
Testimony made clear that the risk is not limited to stranger danger online but can come from systems designed to mimic warmth and understanding. When a machine claims authority, like posing as a therapist, it can steer vulnerable kids away from real help. That combination of mimicry and manipulation is exactly what lawmakers aim to outlaw for minors.
The GUARD Act targets companion chatbots aimed at those 17 and under, bans chatbots from pushing explicit material to minors, and requires clear disclosures that the agent is not human. Supporters argue these are basic consumer protections: kids deserve systems that cannot groom, instruct in self-harm, or pretend to be people. The bill is framed as a proportional response to documented cases of harm.
Hawley did not mince words about the role of corporate incentives. “I mean, it is the worst kind of grooming,” he said at the hearing. “If that was a thing done by a human, the human would be in jail. We would call that sexual grooming.” Those lines underscored a call for accountability when products act like predators.
Lawmakers noted a wave of lobbying from tech companies pushing back at the last minute, yet the committee voted 22-0 to move the measure forward. That unanimous passage gave the bill momentum but left open how quickly the full Senate would act. Political leaders were urged to bring the measure to the floor so families could see concrete action.
Republican sponsors framed the fight as protecting families against an industry that profits wildly while denying responsibility. “No amount of profit justifies the deliberate taking of a child’s well-being, and these companies know very well that this is going on,” one senator said, calling for the Senate to act before more tragedies occur. “This isn’t theoretical. This isn’t about an esoteric problem,” he added. “These are real parents with real children who are basically being extorted by chatbots.”
Graduate Student, wife, engaged political and legal writer.