AI insurance startup AIUC raises $40M to back agent standards and liability frameworks

AI insurance startup AIUC announced a $40M Series A led by Ribbit Capital and First Harmonic, with the round highlighting a growing focus on liability and trust as binding constraints on AI adoption. The company’s work centers on AIUC-1, a standard for agent security, safety, and reliability that its CEO Rune Kvist describes as a framework for auditing autonomous systems for risks such as jailbreaks, hallucinations, and data leaks.

AIUC’s customer list includes Cursor, Harvey, Lovable, and ElevenLabs, indicating early traction among companies shipping agentic products. Kvist, who was among Anthropic’s first hires, framed the Series A as a validation of the hypothesis that risk—not capability—will limit AI’s growth as systems become more autonomous. He noted that the moment for this thesis has arrived with the emergence of advanced agent platforms such as Mythos and Fable.

The company’s approach combines standards development with insurance underwriting, positioning both as infrastructure for deploying autonomous systems. Kvist argued that standards may need to update quarterly rather than on decadal cycles to keep pace with rapid advances in agent capabilities. He also emphasized the need for insurance products tailored to AI risks, including cybersecurity, child safety, and biological threats, and cited the Air Canada chatbot case as an early example of how AI failures are clarifying legal liability.

Kvist discussed the broader trust gap between frontier labs and governments, suggesting that enterprises face an impossible mandate: adopt AI quickly while ensuring nothing goes wrong. He proposed certifications for AI engineers and highlighted the role of red teaming, mechanistic interpretability, and monitoring in building trustworthy systems. The company’s roadmap spans agents, frontier models, robotics, and universal red teaming, with an emphasis on making autonomous systems insurable and legally accountable.