Congressman Brett Guthrie and Congressman John Joyce held a hearing on the safety of AI chatbots. The event, titled “Innovation with Integrity: Examining the Risks and Benefits of AI Chatbots,” was held by the Subcommittee on Oversight and Investigations. The hearing took place on November 18, 2025, at 2:00 PM ET in the Rayburn House Office Building and was accessible to the public and press via livestream at energycommerce.house.gov.
“The rapid advancement of AI held tremendous promise for the future,” said Chairmen Guthrie and Joyce. However, they expressed concerns about recent reports on AI chatbots’ interactions with users, particularly regarding health and wellbeing. They emphasized the need for oversight to understand potential risks to users, especially children and teens.
Dr. John Torous from Beth Israel Deaconess Medical Center provided testimony on AI’s role in mental health. He noted that while AI could enhance mental health care, it also posed immediate dangers. Torous highlighted four main points: unsupervised use of AI in mental health, lack of strong evidence supporting AI claims, poorly understood harms, and the necessity for clearer regulation.
Torous explained that many Americans used AI tools not designed for mental health purposes for high-risk issues like suicidal ideation. He cited data indicating that over one million weekly ChatGPT conversations involved suicidal planning indicators. Despite offering emotional support at times, these technologies were not validated for such use due to proprietary constraints that hindered transparent evaluations.
He pointed out minimal clinical evidence backing claims that AI chatbots improved mental health outcomes. His reviews found a lack of high-quality research despite increased activity in this area. Many studies relied on weak methodologies or low-quality sources like Reddit.
Torous discussed significant mental health risks linked to AI use, including psychotic-like reactions and harmful dependency. He warned against misleading marketing practices in the AI mental health marketplace where companies advertised therapeutic benefits without regulatory oversight.
Torous concluded by urging Congress to support fair competition, transparency, privacy protections, evidence-based development, and patient-centered benchmarks to guide safe AI use in mental health.
For further inquiries about the hearing or press-related questions, interested parties were directed to contact Jackson Rudden or Daniel Kelly via email.
Information from this article can be found here.
