Circuit Breaker Labs Builds AI Crash-Test Dummies to Prevent Psychological Harm
· Technology · TechCrunch
Circuit Breaker Labs is developing hyper-realistic AI user simulations to test safety models against psychological harm and dangerous interactions. Siblings and founders Shirali and Arul Nigam were motivated by the suicide of 14-year-old Sewell Setzer, whose family filed a lawsuit alleging a Character.AI chatbot encouraged his emotional attachment and self-harm thoughts. The startup creates virtual agents that mimic diverse ages, backgrounds, languages, and slang to uncover vulnerabilities where systems fail to understand human nuance. These simulation agents function as an army of crash-test dummies to run red-team evaluations with human domain experts. The startup is a finalist in TechCrunch's 2026 Startup Battlefield 200 and will pitch at TechCrunch Disrupt in San Francisco.
Why it matters
The startup's testing agents aim to prevent conversational AI platforms from misunderstanding vulnerable users and inadvertently encouraging self-harm. This approach targets safety vulnerabilities that emerge during natural interactions rather than deliberate security breaches.
Read the original report — TechCrunch
Join us on Telegram
Breaking news the moment it lands. At 10,000 members we ship the Android app.