AI Safety Startup Fights Psychological Harm

AI interactions have unfortunately become psychologically dangerous for some, leading to tragic outcomes. This has resulted in wrongful death lawsuits against AI companies, alleging their bots played a role in user suicides.

Circuit Breaker Labs, a finalist in TechCrunch's Startup Battlefield 200, is dedicated to enhancing AI safety across diverse languages and cultures. Siblings Shirali and Arul Nigam founded the company, inspired by a case where a teenager allegedly developed a harmful attachment to an AI chatbot.

The startup employs AI agents, described as "crash-test dummies," to simulate users of various ages, backgrounds, and linguistic styles. These simulations rigorously test AI models for their ability to detect and appropriately respond to psychologically harmful interactions, especially those involving nuanced language or slang.

Circuit Breaker Labs conducts extensive "red-team" tests, simulating tens of thousands to hundreds of thousands of interactions daily. Their goal is to ensure AI systems can safely handle risky conversations and prevent users from falling into "AI psychosis" or developing unhealthy parasocial relationships with bots.

While currently operating as a safety testing lab for high-risk AI applications like mental health support apps, Circuit Breaker Labs aims to build trust in AI. They believe making AI safer is the key to overcoming public skepticism and preventing a regressive ban on potentially valuable technology. The company will present at TechCrunch Disrupt in San Francisco from October 13-15, 2026.