A bipartisan Senate bill that would ban minors from using AI companion chatbots, force platforms to verify users’ ages, and create new civil and criminal penalties for chatbots that encourage self-harm cleared a major hurdle this year, and it is now awaiting a vote by the full Senate. The Guidelines for User Age-verification and Responsible Dialogue Act — known as the GUARD Act — represents the most sweeping federal attempt yet to regulate the fast-growing category of AI apps designed to simulate friendship, romance, and emotional support.
What the Bill Actually Does
The GUARD Act, formally S. 3062, targets what lawmakers call AI companions: chatbots built to simulate a sustained interpersonal or emotional relationship with a user, as opposed to general-purpose assistants. Under the bill, providers of these companion apps would have to implement "reasonable age verification measures" across all user accounts, and if a user is determined to be under 18, the platform could not allow that account to access the companion features at all. Chatbots would also be required to clearly and repeatedly disclose that users are talking to a non-human system that lacks professional credentials — meaning a bot could not imply it is a licensed therapist, doctor, or attorney, or that it is providing medical, legal, financial, or psychological services.
The bill goes further than disclosure rules. It would make it unlawful to offer an AI chatbot to the public when a provider knows, or recklessly disregards, that the chatbot encourages suicide, self-injury, homicide, or imminent physical or sexual violence, or when it engages minors in sexually explicit conversation. Violations could trigger civil actions brought by the U.S. Attorney General or state attorneys general, with civil penalties of up to $250,000 per violation, and companies that cross the bill’s bright lines on self-harm or sexual content involving minors could face criminal liability with fines up to $250,000 per offense.
Unanimous Committee Vote
The Senate Judiciary Committee advanced the GUARD Act on April 30, 2026, in a vote reported as unanimous, according to coverage from the policy-tracking outlet GlobalPolicyWatch. The bill was introduced in October 2025 by Senators Josh Hawley, a Missouri Republican, and Richard Blumenthal, a Connecticut Democrat, with roughly seventeen senators from both parties signing on as cosponsors — an unusually broad coalition for tech legislation in a chamber where bills on content moderation and platform liability have historically stalled. At the markup session, Hawley reportedly thanked what he described as the "brave parents whose children were abused by these AI company chatbots," a reference to a string of lawsuits and news reports describing chatbots that allegedly groomed, manipulated, or encouraged children toward self-harm.
The Backdrop: A Wave of Family Lawsuits
The push for federal legislation did not emerge in a vacuum. Over the past two years, families have filed wrongful-death and negligence suits against major AI companies, including Character Technologies and OpenAI, alleging that companion chatbots contributed to their children’s suicides or severe psychological harm. Those cases, combined with a Federal Trade Commission inquiry launched in September 2025 into how seven major companies — including Alphabet, Meta, Snap, OpenAI, Character Technologies, and xAI — test and monitor the effects of companion bots on children, have kept pressure on Congress to act where company self-policing has fallen short.
Not Everyone Is on Board
The bill’s rapid, bipartisan advance has not silenced critics. Civil-liberties and free-speech organizations, including the think tank R Street, have argued publicly that the GUARD Act raises First Amendment concerns and could be difficult to implement without sweeping, privacy-invasive age-verification systems — such as facial recognition or government ID checks — that sweep up adult users along with minors. Supporters of stronger verification, including parents’ advocacy groups like the Tech Oversight Project, have applauded the committee’s action but want the bill to resist industry lobbying to weaken verification standards to merely "commercially reasonable" methods rather than firmer requirements.
A Patchwork Already Exists
Even before the GUARD Act reaches the Senate floor, states have been moving on their own. California’s SB 243, signed by Governor Gavin Newsom, took effect January 1, 2026, and requires AI companion operators to verify ages, disclose that conversations are AI-generated, restrict romantic or sexual interactions with minors, and maintain crisis protocols for users expressing suicidal ideation. Washington, Oregon, Idaho, Nebraska, Iowa, Georgia, Connecticut, Colorado, and Hawaii have all signed companion-chatbot laws of their own this year, and lawmakers have introduced close to 100 chatbot-related bills across 34 states, creating a compliance patchwork that has itself become an argument for federal preemption-style legislation like the GUARD Act.
What’s Next
The bill now needs a floor vote in the full Senate, where its fate is less certain than in committee, and companion legislation has been introduced in the House. Lawmakers could still amend verification standards before a floor vote, and legal challenges are widely expected if the bill becomes law, given the First Amendment objections already raised. For families who have sued AI companies, and for the companies themselves, the GUARD Act’s progress signals that Washington — not just state capitals — now sees AI companion chatbots and child safety as an urgent, bipartisan priority, even if exactly how to regulate the technology remains contested.
Photo: cottonbro studio / PEXELS via Pexels
