Corgi Is Worth $1.3 Billion. Whether It Has Real Insurance Is a Different Question.
Four months. A $670 million valuation step-up. That's the bet Corgi Insurance just sold to investors.
The San Francisco startup closed a $160 million Series B at a $1.3 billion valuation on May 6, 2026, four months after a $108 million Series A at $630 million — a gap that implies investors think the competitive window for AI-native insurance is closing fast. Corgi sells policies covering AI-specific failures: biased algorithms, inaccurate content, training data misuse, adversarial model attacks, synthetic media, and autonomous system failures. Traditional carriers have begun excluding AI-related risks from their policies, leaving a gap Corgi is built to fill. Deel and Artisan are customers, and Forbes reports Corgi also insures Eragon. Forbes reports Corgi claims annualized revenue around $100 million, with over 40,000 active startup customers across 49 states and churn below 1%. Founded in 2024 by Nico Laqua and Emily Yuan after Y Combinator's Summer 2024 batch, TCV led the Series B. It is expanding into trucking.
What will determine whether that valuation holds is also what Corgi cannot yet prove: what happens when a major claim lands. AI harms introduce causation problems standard insurance law hasn't resolved. When an autonomous system causes harm, the causal chain typically runs through multiple parties — the model vendor that built the underlying system, the company that deployed it, the system integrator that connected it to their workflow, and the end user who fed it inputs. If an AI hiring tool discriminates, a content tool generates defamatory material, or an agent gets tricked into authorizing a fraudulent payment, all four links in that chain can plausibly be held liable. Insurance policies written for each party may or may not cover the same incident depending on how policy language defines the triggering event — and "autonomous system failure" is a phrase that means different things to the company that wrote the policy and the lawyers who will dispute it.
Corgi's policies explicitly name adversarial model attacks and autonomous system failures as covered scenarios. The company says its Hammurabi risk engine produces a policy quote in under five minutes and claims 95% recall accuracy at scale — a vendor claim describing pricing accuracy, not claims-processing accuracy. The insurance question isn't whether Corgi can price risk accurately on the way in. It's whether it can pay on the way out, when the parties involved are arguing about causation, policy language, and which link in the chain the policy was actually covering.
Regulatory approval came in July 2025. The company is twelve months old as a licensed carrier. The cafe loses money, Laqua told Forbes, but he considers it worth it for community and recruiting. A 22-year-old founder working from the cafe put it simply when she described the experience to Forbes: "They understand what startup speed is." Corgi has a corgi mascot, a named underwriting engine, paying customers, and a valuation that prices in the assumption that AI liability insurance will be a durable category — and that it will be the company that wins it. Whether its policies hold when a court or arbitrator reads them, and when the underlying model vendor's own liability terms are also in the room, is what the next 12 months of claims history will begin to answer.