Circuit Breaker Labs, founded by siblings Shirali and Arul Nigam, is building a testing platform that simulates users of all ages, languages and slang to probe AI chatbots for psychologically dangerous replies. The company will pitch its product at TechCrunch Disrupt in San Francisco from October 13‑15, 2026. It runs tens of thousands of simulated conversations each day and scores models on how well they handle risky interactions.
Why it matters
If the testing catches harmful responses, mental‑health apps and other chatbot services can avoid prompting suicidal or distressing behavior in vulnerable users.
