Senate’s GUARD Act Narrows Focus to AI Companions, Raising Privacy Concerns

The Senate’s GUARD Act has been narrowed from an earlier, broader approach to focus on “AI companions,” but the fight over the bill is now centered on privacy and how it would work in practice. The Electronic Frontier Foundation, or EFF, said in a May 8 analysis that the revised measure still relies on identity-linked age verification and uses definitions that could chill access to some AI tools and discourage development.

The bill, S.3062 in the 119th Congress, is formally titled the Guidelines for User Age-verification and Responsible Dialogue Act of 2025. It was introduced Oct. 28, 2025, with Sen. Josh Hawley as sponsor and bipartisan cosponsors including Sen. Richard Blumenthal. The Senate Judiciary Committee considered it April 30 and committee materials, along with independent coverage, indicate it advanced that day. The bill would bar minors from AI companions and require companies to verify users’ ages.

According to the introduced text now posted on Congress.gov, the measure applies to an “AI companion,” defined as a type of AI chatbot designed to simulate interpersonal or emotional interaction, friendship, companionship or therapeutic communication. Covered companies would have to create user accounts and use a “reasonable age verification” process, with periodic re-verification. The bill also would require chatbots to disclose that they are not human at the start of a conversation and every 30 minutes after that. If age verification shows a user is under 18, the company could not allow that person to access an AI companion. Enforcement would be shared by the U.S. attorney general and state attorneys general.

EFF said lawmakers did respond to criticism by narrowing the bill from language that could have swept in a much wider range of AI-powered chat and search tools. But Joe Mullin, the author of the group’s May 8 analysis, argued that the revised version still creates major problems. EFF said the bill points companies toward age checks tied to real-world identity, including methods linked to financial records or age-verified mobile operating system and app store accounts. Privacy groups and some industry critics say that kind of online age assurance can create surveillance risks and access barriers for people who lack government ID, bank accounts or stable digital identity tools. EFF also said the new definition of AI companion may still be vague enough to reach some emotionally responsive systems, including certain customer-service chat tools, while potential liability could push smaller developers to block minors entirely or disable features.

There is also an unusual transparency problem around the bill’s latest version. EFF and some other reporting say the revised measure increased penalties to $250,000 per violation. But the public text currently posted on Congress.gov still shows the introduced version, which sets civil penalties of up to $100,000 per violation and some criminal fines of up to $100,000 per offense. As of May 8, a committee-reported or amended text confirming the higher figure did not appear to be publicly available on Congress.gov or the Judiciary Committee’s site. That leaves an important gap between reported committee changes and the authoritative text the public can independently review.

Congress is acting amid a broader 2025-2026 push in Washington and in statehouses to regulate AI companion chatbots after reported harms involving minors, including lawsuits alleging manipulative or dangerous interactions. Supporters say stricter rules are needed to protect children. In announcing a House companion bill on April 30, Rep. Valerie Foushee, D-N.C., said, “People under the age of 18 should not be able to interact with AI chatbots.” That House measure was introduced by Foushee and Rep. Blake Moore, R-Utah. As the debate moves forward, child-safety advocates have praised the effort, while privacy and digital rights groups argue the details of age verification and the bill’s scope remain unresolved.