Underwriting Superintelligence: Backing Agents you can Sue — Rune Kvist, AIUC
Article image or reusable cover for Latent Space
AIUC, founded by Rune Kvist from Anthropic, announced a $40 million funding round to solve a growing problem: who is liable when AI agents make mistakes?
The company develops AIUC-1, a standard for testing and insuring AI systems against jailbreaks, hallucinations, and data leaks. Kvist argues that trust—not AI capability—will become the biggest constraint on deploying AI in the real world. Companies like Cursor, Harvey, and ElevenLabs face increasingly difficult questions about responsibility when autonomous systems cause harm, and AIUC is building testing, standards, and insurance infrastructure to enable this.
Vibekollen prepared this summary with AI from the original publication. The content belongs to Latent Space.
