Circuit Breaker Labs hopes to make AI safer for your kids (and you)

Circuit Breaker Labs is developing what it describes as an army of AI crash-test dummies: agents that mimic people of different ages, backgrounds, languages, and cultures to test whether models can detect dangerous, psychologically harmful interactions. The company is one of TechCrunch's 2026 Startup Battlefield 200 finalists and will pitch at TechCrunch Disrupt, held at Moscone West in San Francisco from October 13-15.

Siblings Shirali and Arul Nigam founded the startup. Arul, the CTO, said they were motivated by Sewell Setzer, the 14-year-old who developed an emotional attachment to a Character.AI chatbot and confessed thoughts of harming himself before dying by suicide. His parents alleged in a 2024 lawsuit that the chatbot encouraged him. Arul said the bot may not have understood what words like "I want to be with you" implied.

The company says it works with human domain experts to build hyper-realistic user simulations for red-team tests designed to uncover weaknesses, reflecting real speech patterns, slang, coded language, and typos. It runs tens of thousands to hundreds of thousands of simulated interactions per day and uses a proprietary scoring method to produce auditable, explainable scores.

Shirali, the CEO, said a six-year-old girl, a 45-year-old man, first- versus second-language English speakers, or different slang can trip up a model. Models handle standard speech patterns well, she said, but people do not talk that way, and misunderstanding nuance or slang can go badly.

Circuit Breaker Labs operates as an AI safety testing lab for high-risk applications such as AI coaching, journaling, or other mental health support apps. Arul declined to name its marquee customers. It has a working product but is very early, with five employees including the Nigam siblings. The platform could eventually be applied to apps where someone may fall into an "AI psychosis" hole or develop a parasocial relationship with a chatbot, such as AI "co-worker" agents whose responses vary between interactions.

Arul said skepticism of AI is healthy but banning a potentially valuable tool over safety concerns would be "regressive." He said the company believes the answer to those fears is making AI safer: "We want to help build that trust for people."