US Senate Judiciary tees up AI chatbot, companion safety debate

Momentum is building for U.S. Congress to address children's online safety issues stemming from the rampant proliferation of artificial intelligence chatbots. The proposed Guidelines for User Age-verification and Responsible Dialogue Act, a bill dedicated to children's protections against chatbots and companions, is now available for full Senate consideration following unanimous approval by the Senate Committee on the Judiciary.

The bipartisan GUARD Act, led by Sen. Josh Hawley, R-Mo., aims to limit children's access to harmful and explicit content via chatbot interactions that could negatively impact their mental health or well being. The bill proposes requirements for privacy-preserving age verification to help limit access while prohibiting users under age 18 from interacting with AI companions.

"We're often told that this new dawning age of artificial intelligence is going to be a great age that will strengthen families and workers. I would just say that's a choice, not an inevitability," Hawley said during the Senate Judiciary markup session 30 April. "(The GUARD Act) is a common sense, targeted, tailored effort to protect our kids and protect our families."

The bill also contains requirements for providers to disclose to users before and during interactions that they are not interacting with a human or a licensed professional.

Enforcement lies with the U.S. and state attorneys general, and includes the potential for up to USD250,000 in civil penalties per violation. U.S. attorney general will receive rulemaking authority while states can apply their own laws to violations if they are "at least as protective of users of artificial intelligence chatbots as this Act."

Hawley made clear the bill does not ban chatbots and companions altogether, noting providers are still encouraged to offer age-appropriate experiences for learning and leisure purposes alike. The clarification seeks to highlight balance between the need for enhanced children's safety and U.S. AI innovation goals.

He characterized the bill as taking a "very modest" approach to chatbots, noting he and cosponsors "could've gone a lot broader, and there were times I thought we should have."

If it is passed and signed by the president, the GUARD Act would take effect 180 days after enactment.

Lawmakers aligned

Bipartisanship exists in the Senate on children's online safety improvements, evidenced by ongoing work toward passing the Kids Online Safety Act and the Children and Teens' Online Privacy Protection Act. The GUARD Act, along with the newly introduced Children’s Health, Advancement, Trust, Boundaries, and Oversight in Technology Act, represent the next wave of Senate consensus.

Lawmakers are also cognizant of their shortcomings with regulating social media platforms, which is proving to be motivation to expedite work on chatbot regulations to match the speed at which AI is developing.

"AI is even more powerful, developing even more quickly and we have even less time to do an effective job at reining it in," Sen. Chris Coons, D-Del. said. "On the positive end, it hold enormous promise to do incredible things. But on the negative end, it is very dark and it is very dangerous."

On the same day it moved in the Senate, the GUARD Act's House companion was introduced. Reps. Blake Moore, R-Utah, and Valerie Foushee, D-N.C., will lead the House consideration of the bill.

In a statement, Moore said the GUARD Act raises "a critical step to draw lines in the sand with Big Tech" and will ultimately steer minors toward more in-person interactions rather than send them toward the "unaccountable black hole of frontier technology."

KOSA and COPPA 2.0 are among the the Senate-approved children's bills that have stalled in House in recent years after lower chamber amendments that jeopardized bipartisanship. Foushee wants to avoid the same fate for the GUARD Act.

"Our children are our top priority, and we have a responsibility to implement proper safeguards to ensure they are not being negatively impacted by AI," She said in her statement. "I'm proud to introduce the bipartisan and bicameral GUARD Act with Congressman Moore, and I will continue to advocate for further safeguards that protect our communities from the harms and risks associated with AI."