Early Anthropic hire, former METR COO have found a way to rein in rogue AI agents — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- AIUC announced a $40 million Series A led by Ribbit Capital with First Harmonic participating, on top of a prior $15 million seed round from Nat Friedman's NFDG, Emergence, Terrain, and Anthropic co-founder Ben Mann — bringing total funding to $55 million.
- AIUC (Artificial Intelligence Underwriting Company) was founded by brothers-in-law Rune Kvist, an early Anthropic employee, and Rajiv Dattani, the former COO of AI safety research org METR (2024-2025), and already counts Cursor, Lovable, Harvey, and ElevenLabs as customers.
- AIUC built the AIUC-1 standard modeled after the SOC 2 cybersecurity framework, shaped by a consortium of roughly 250 enterprise security and risk leaders who buy AI agents.
- AIUC runs about 5,000 tests per agent covering jailbreaks, hallucinations, and data leaks, producing a ~100-page audit report — with AI agents running the tests and AI analyzing the data, and humans only verifying the final audit.
- Rajiv Dattani remains a board member at his old firm METR, which has focused on frontier-model performance testing and was one of the independent orgs OpenAI used to investigate its Hugging Face incident.
- Dario Amodei recently called for frontier labs to use embedded third-party evaluators and named METR as a possibility, echoing AIUC's pitch of independent AI safety certification for enterprise buyers.
Why it matters: Banks, hospitals, governments, and militaries are declining to deploy AI agents not because models are too dumb but because nobody can guarantee what the agents will do, per AIUC — a SOC 2-style certification gives procurement teams a third-party signal to greenlight agent purchases. The $40M Series A led by Ribbit Capital bets that AI auditing becomes a purchasing prerequisite for enterprises before regulators mandate one.
Ask SkimNews
