A $40 Million Plan to Keep AI Helpers From Going Off Track
Artificial Intelligence Underwriting Company (AIUC) has raised $40 million in a Series A round led by Ribbit Capital. The round brings its total funding to $55 million, with First Harmonic participating, according to the disclosure. TechCrunch
AIUC was started by Rune Kvist, an early employee at Anthropic, and Rajiv Dattani, who was chief operating officer of AI safety group METR from 2024 to 2025 and is still a board member there. The new funding follows a $15 million seed round. Seed backers included Nat Friedman through NFDG, Emergence, Terrain, and Anthropic co-founder Ben Mann.
AIUC lists Cursor, Lovable, Harvey, and ElevenLabs as customers. It built a standard called AIUC-1 for testing AI agents.
The broader context here is simple. A quiz for the AI model is not enough anymore. Once an AI helper can click buttons, remember past chats, ask other AIs for help, and work with real company data, mistakes are about what it does, not just what it says. Think of it like a new assistant with keys to the office. You need clear rules about which doors it can open.
Looking at what this means for teams building these tools, the name underwriting fits. It means checking safety over time, not once. Engineers will ask practical questions. Where do checks happen before software goes live. What records are saved. How rules are fixed when broken. How each AI action is tracked across other apps. AIUC wants its AIUC-1 checklist to work across companies, which could make it cheaper to use if buyers agree.
In my view, the customer list matters most. Coding helpers, legal tools, and voice and video tools let AI take actions that count at work. Those jobs have strict rules about privacy and review. If safety checks become a normal buying requirement there, others will follow. That would let more useful AI reach workers with fewer surprises.
