Lisbon startup raises seed funding for AI agent insurance

A Lisbon-based startup has raised seed funding to develop insurance products for autonomous AI agents.

The startup appears to be among the first to focus on underwriting risk for AI agents operating with minimal human oversight. Its policies aim to cover financial losses that could arise if agents make errors—such as incorrect data processing, flawed execution, or misrouted operations—that might impact a company’s operations. The company has reportedly issued policies to a small number of pilot customers, though details about their identities or use cases remain undisclosed.

The challenge is straightforward in concept but difficult in execution. Insurers typically rely on historical loss data to price risk, yet few AI agents today operate at a scale where losses are measurable. Without established benchmarks, the startup’s approach may involve modeling risk through simulations or controlled testing, potentially drawing from methods used in other emerging-risk insurance sectors. However, applying these techniques to AI agents remains unproven.

Recent developments suggest this space is evolving rapidly. Earlier coverage highlighted startups monitoring AI agent behavior within companies, while others are expanding tools to track agents across multiple platforms. Meanwhile, some startups are deploying agents in live business environments, where errors could have financial consequences. If this trend continues, demand for insurance could emerge as a critical enabler—or bottleneck—for broader adoption.

The core tension in this model is timing. Most AI agents today remain in pilot phases, where mistakes are expected and tolerated. Insurers generally avoid covering experimental technology, meaning early adopters of such policies might be startups seeking insurance to meet enterprise requirements. This dynamic could create a feedback loop: insurers need data to price policies, but startups may struggle to obtain coverage without it.

To navigate this, the startup may be starting with narrow, well-defined tasks where losses are easier to quantify and agent decisions can be audited. Over time, the ambition would likely expand into higher-stakes domains like supply chain automation or financial services, where the potential risks—and need for insurance—would be greater.

The seed round indicates investor confidence in the long-term necessity of this layer in the AI agent ecosystem. The lead investor has previously backed infrastructure plays in emerging tech, while the other specializes in impact-driven startups. Their bet hinges on the assumption that insurance will become a prerequisite for AI agent adoption, and that early movers could establish a durable position.

What happens next will depend on how quickly AI agents transition from pilots to production. If startups begin landing large enterprise deals, demand for insurance could accelerate. Regulatory requirements for financial safeguards could also drive adoption. The startup’s challenge will be scaling its underwriting models ahead of the market while avoiding early losses that might deter reinsurers.

For now, the company operates in a space with few competitors, though that may change. Larger insurers have shown interest in AI-related risk products, and if this model gains traction, they could enter the market. The real test will be whether the startup can expand beyond insuring early-stage companies and begin serving the enterprises that may eventually deploy AI agents at scale. Until then, it remains a solution for a market that doesn’t yet exist—but one that could soon become essential.

Sources: sifted.eu