AIUC Lands $40M to Expand AI Model Auditing Beyond Agent Layer

Artificial Intelligence Underwriting Company announced a $40 million funding round aimed at extending its certification work to frontier AI models, marking a significant expansion of its scope. The startup has previously focused on auditing AI agents that enterprises build on top of foundation models, but the new capital will fund a shift toward evaluating the models themselves.

AIUC's core argument is that risk uncertainty, rather than technological limitations, now represents the primary obstacle to broader AI adoption. Co-founder and Chief Executive Rune Kvist noted that many enterprises maintain inventories of agents that advanced through pilot programs but became stalled during security review phases. He identified proof of security and reliability as the critical bottleneck preventing deployment.

How the certification works

AIUC's certification standard, designated AIUC-1, subjects each agent to approximately 5,000 risk and attack scenarios customized for the deploying organization's industry vertical. Testing covers jailbreaks, hallucinations, data leaks and other failure modes. Agents undergo independent audits and must be recertified quarterly as attack methodologies evolve. According to TechCrunch, much of the testing relies on AI agents themselves, with human reviewers validating the final audit results.

Current certified deployments

Voice AI company Eleven Labs Inc. leveraged its AIUC certification in February to obtain what it characterized as the first insurance policy of its kind for AI agents, including protection against scenarios where a voice agent delivers inaccurate customer information. Other holders of the certification include Cursor developer Anysphere Inc., legal AI company Harvey AI Corp., KPMG LLP, Lovable Labs Inc., UiPath Inc., and customer service software maker Fin, which was formerly known as Intercom Inc.

A consortium comprising more than 250 security and risk executives from Fortune 1000 companies participates in shaping the AIUC-1 standard and actively promotes its adoption within their respective organizations.

Founders and their backgrounds

Kvist served as Anthropic PBC's inaugural product hire before co-founding AIUC. His co-founder, Rajiv Dattani, previously worked as a partner within McKinsey & Co.'s insurance division and later held the chief operating officer position at METR, a nonprofit focused on AI model evaluation.

Dattani drew a historical parallel to underscore the company's mission: "When electricity was burning down houses, the insurers paying the bill funded Underwriters Laboratories to test and certify products. To this day, the UL mark is on most light bulbs across America. AI needs the same combination of standards, testing and insurance."

Expansion strategy

AIUC initially concentrated on agent applications because security reviews emerged as the first major deployment bottleneck. The company now observes similar pressures building around frontier models themselves. The funding will enable AIUC to expand its audit and insurance capabilities to encompass that layer, contributing to what the company describes as a comprehensive oversight ecosystem for the most sophisticated AI systems.

Funding details

Ribbit Capital led the Series A round with support from First Harmonic. The company previously raised $15 million in seed funding led by Nat Friedman at NFDG when it launched in July 2025. Total investor commitments now reach $55 million.

Micky Malka, founder and managing partner of Ribbit Capital, highlighted his firm's decade-long track record backing financial services companies operating in sectors where trust supersedes all other considerations. He characterized AI as following an analogous trajectory but at an accelerated pace, noting that "it is moving faster than the systems companies use to evaluate it."