Reps. Valerie Foushee, Mike Thompson, Gun Violence Prevention Task Force Press AI Executives on Chatbot Safety

WASHINGTON, DC (June 4, 2026) - Today, Congresswoman Valerie Foushee (NC-04), Congressman Mike Thompson (CA-04), Chairman of the House Gun Violence Prevention Task Force, and members of the task force sent a letter to OpenAI and Google DeepMind urging the companies to address the risks posed by online chatbots and requesting information on the steps they are taking to protect consumers and public safety.

"Given gun violence is the leading cause of death among children and teens, we urge you to acknowledge the serious risks your AI systems may be exacerbating and to take immediate action to address them," the lawmakers wrote. "We are particularly alarmed by reports that individuals have consulted your AI chatbots regarding carrying out criminal acts, committing suicide, or confided thoughts incompatible with the safety and welfare of themselves or others."

"In these interactions, chatbots reinforced, rather than dissuaded, real-world harm, including mass shootings, wrongful deaths, and suicide," the lawmakers continued. "The language generated by these chatbots has undoubtably led to serious harm. By providing information and strategies to shooters, they have contributed to preventable loss of life."

"As chatbots become more pertinent to society and individuals' lives, it is time to address the risks of chatbots providing hazardous information and AI-induced psychosis," the lawmakers concluded. "While we continue to hope that current efforts and future models ensure user safety, we must request a staff-level congressional briefing on these matters, along with written responses."

The letter follows reports and legal filings alleging that AI chatbots were used in connection with violent and harmful acts in British Columbia, at Florida State University, and at Miami International Airport. This week, Florida's State Attorney General filed the first lawsuit by a state against OpenAI and CEO Sam Altman, alleging that the company's chatbot failed to adequately protect users.

The letter demands that OpenAI and Google DeepMind provide written responses regarding:

• Whether the companies, their subsidiaries, affiliated entities, or predecessor entities have been subject to any state or federal investigation, regulatory inquiry, enforcement action, fine, penalty, civil judgment, settlement agreement, or consent decree related to allegations of physical harm, mental or emotional harm, wrongful death, homicide, suicide, or facilitation of violence.
• What user data the companies collect from minors, how that data is used to train models, and how a user's age affects chatbot behavior.
• What safeguards the companies tested before releasing products to minors or users with varying mental-health statuses.
• What policies are in place when users discuss harming themselves or others, or request information about deadly weapons, including whether the companies notify law enforcement or mental-health organizations when a user shares plans to harm themselves or others.

The letter was signed by Representatives Valerie Foushee (NC-04), Mike Thompson (CA-04), Gabe Amo (RI-01), Yassamin Ansari (AZ-03), Don Beyer (VA-08), Julia Brownley (CA-26), Salud Carbajal (CA-24), Judy Chu (CA-28), Angie Craig (MN-02), Jasmine Crockett (TX-30), Danny Davis (IL-07), Madeleine Dean (PA-04), Chris Deluzio (PA-17), Mark DeSaulnier (CA-10), Dwight Evans (PA-03), Maxwell Frost (FL-10), Jahana Hayes (CT-05), Jonathan Jackson (IL-01), Hank Johnson (GA-04), Robin Kelly (IL-02), Tim Kennedy (NY-26), Raja Krishnamoorthi (IL-08), Stephen Lynch (MA-08), Jennifer McClellan (VA-04), Kelly Morrison (MN-03), Mike Quigley (IL-05), Deborah Ross (NC-02), Andrea Salinas (OR-06), Jan Schakowsky (IL-09), Kim Schrier (WA-08), Rashida Tlaib (MI-12), Jill Tokuda (HI-02), Paul Tonko (NY-20), Marc Veasey (TX-33), and Debbie Wasserman Schultz (FL-25).

Read the full letter here.