AIUC raises $40 million to test and certify AI agents for businesses

In short: A new startup called AIUC raised $40 million to help companies check whether their AI agents behave safely and predictably.

What happened

Artificial Intelligence Underwriting Company, or AIUC, announced a $40 million Series A funding round led by Ribbit Capital, with participation from First Harmonic. The company was founded by Rune Kvist, an early employee at Anthropic, and Rajiv Dattani, a former operations leader at the AI safety group METR.

AIUC sells audits for “AI agents,” which are AI tools that can take actions on their own, like searching for information, writing messages, or moving data between apps (think of them as a digital assistant that can click buttons for you). The company says it built a standard called AIUC-1, inspired by SOC 2, a well-known cybersecurity checklist many businesses already use.

To create its standard, AIUC says it gathered input from about 250 security and risk leaders. It then tests an AI agent through around 5,000 scenarios, including “jailbreaks” (ways people try to trick an AI into ignoring rules), “hallucinations” (when an AI confidently makes things up), and data leaks. AIUC produces a detailed report, and the founders say humans review the final audit even if AI helps run the tests.

Why it matters

More companies want to use AI agents, but many worry about surprises, like an agent sharing private data or taking the wrong action. AIUC is betting that independent testing and clear reports will make it easier for businesses, banks, and hospitals to decide which AI tools they can trust.

Source: TechCrunch AI