Klaimee, a Y Combinator-backed insurtech startup, has raised $5.5 million in seed funding to build what it calls the first purpose-built liability coverage for autonomous AI agents. The round was led by FundersClub's Alexander Mittal, with participation from ex/ante, Pioneer Fund, Multimodal Ventures, Kima Ventures, Rebel Fund, Robinhood Ventures, Y Combinator and a group of angel investors.

Who bears the cost when an agent errs

Klaimee was founded by CEO Ines Boutemadja, previously general manager at SafetyWing, and CTO Julien Catonnet, a former software engineer at VC-backed startups. The company's pitch is aimed at a question it argues the market hasn't answered: as businesses hand agents access to systems, money and customer communication, who is on the hook when an agent makes a costly mistake?

Klaimee's product combines an errors-and-omissions-style insurance policy with a certification process meant to price that risk. A proprietary scoring system evaluates an agent across three areas before underwriting coverage: a public data scan, a governance assessment made up of roughly 30 questions, and behavioral testing that runs more than 100 probes for failure modes such as prompt injection, jailbreaking, decision drift, data leakage and biased outputs. The company says it can complete certification, underwriting and the procurement paperwork enterprises require in under 24 hours.

Why it matters

Klaimee's raise is a small deal by dollar amount, but it's a concrete example of a market forming around a problem that's become more urgent as companies push agents from chat assistants into systems with real write access — placing orders, moving money, or editing production data without a human approving every step. As enterprises adopt agents faster than their governance and insurance frameworks can keep up, demand is emerging for third parties to price and absorb that risk, similar to how cyber-insurance grew alongside enterprise cloud adoption. If certification-backed agent insurance catches on, it could become a de facto compliance checkpoint that enterprises require of agent vendors before deployment — turning "is this agent insurable" into a practical gate alongside security and accuracy benchmarks.