AIUC raises $40M as Cursor, Harvey back agent standard

AIUC's pitch rests on a pattern its cofounder traces through history: whenever a powerful technology arrives — electricity around 1900, cars in the 1930s, private nuclear energy in the 1950s — the market creates confidence infrastructure before regulation catches up, pairing standards with insurance to make adoption possible. He points to Waymo as the current example: the cars are widely considered better drivers than humans, yet you still cannot take one to the airport, because the binding constraint is liability and trust rather than capability. AIUC argues the same problem now applies to frontier AI, and that it gets harder as models improve, since more capable agents are more autonomous and expand the risk surface.

The company's answer is AIUC-1, a standard for agent security, safety, and reliability built around three design choices: consolidate the concerns of Fortune 1000 security leaders into one framework, ground it in technical testing rather than paperwork, and refresh it quarterly because AI moves too fast for standards that update on a decade cycle. A consortium of risk leaders from real banks, hospitals, and critical infrastructure meets twice a quarter to shape what the standard covers.

The stakes for AIUC hinge on whether it can keep that quarterly refresh meaningful and whether insurers — Lloyd's of London is cited as an example — will actually price and cover agent failures. Its customers, companies like Cursor, Harvey, Lovable, and ElevenLabs, have little trouble selling a pilot; the test is whether certification and insurance carry them through a bank's or hospital's risk review and into wall-to-wall deployment.