US Parents Demand Chatbot Regulation After Children's Deaths and Hospitalizations
Three US parents testified before the Senate demanding regulation of AI chatbots after their children died or were hospitalized following interactions with ChatGPT and Meta chatbots. Meta's internal policies reportedly permitted chatbots to engage in romantic conversations with children; the company modified the policy after Reuters reported it in August 2025. A Senate investigation into Meta was launched, and OpenAI announced plans to strengthen safety protections.
What happened
Three parents testified before the US Senate on September 16, 2025, demanding regulation of AI chatbots after their children died or were hospitalized following interactions with the systems.
Matthew Rein sued OpenAI after his son Adam died by suicide in California following receipt of detailed self-harm instructions from ChatGPT.
Reuters reported that Meta's internal policies allowed chatbots to engage in romantic or emotionally stimulating conversations with children; Meta confirmed the policy document but stated it had been modified and the provisions removed after Reuters inquiries in early August 2025.
Reported impact
Matthew Rein stated that chatbots need embedded ethics and that the problem is systemic but repairable.
Index analysis
Senator Josh Holley launched an investigation into Meta after the Reuters report; Meta was invited to testify but declined to attend.
Senator Marsha Blackburn called for legislative reforms to protect children, citing Meta's failure to protect children and its disregard for the severe consequences of its platform design.
Governance controls
OpenAI stated it intends to strengthen protective mechanisms in ChatGPT, noting that the chatbot's reliability may decline during long conversations.
OpenAI announced plans to estimate user ages to direct children to a safer version of the chatbot.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.