17 Feb, 2026

As insurers retreat from AI risk, one startup plans to fill the gap

The insurance industry is at a crossroads as traditional carriers back away from covering risks posed by AI, leaving companies exposed as adoption of the technology surges.

As some brokers and underwriters shy away, Testudo, a startup with backing from Lloyds of London, is aiming to bridge the gap by not only providing coverage for AI risk, but also helping companies to understand and mitigate risk in the areas where they might be most vulnerable.

Insurers are inevitably going to be affected by GenAI risk whether it be in the auto space, which has seen a rise in self-driving cars, or in segments like professional indemnity insurance services, according to Testudo CEO George Lewin-Smith.

"Ultimately we think every line is going to have some sort of like impact when it comes to AI risk because it's just been deployed so broadly and deeply," Lewin-Smith said in an interview. "That's why we're keen to being here early in the market and start underwriting."

The evolution of AI risk

SNL Image

Testudo CEO George Lewin-Smith.

Source: Testudo.

Major carriers including American International Group Inc., W. R. Berkley Corp. and Great American Insurance Co. have filed requests for regulatory approval to constrain their liability for claims stemming from use of AI systems such as chatbots, among other automated products, according to a media report from the Financial Times.

AI-related claims can come in many different forms and have the potential to reach billions of dollars. For example, artificial intelligence company Anthropic PBC signed a $1.5 billion settlement in September 2025 with authors who had sued it for violating copyright laws by illegally pirating a half-million books to train its chatbot.

AI technology has also been used for nefarious purposes. In 2024, a deepfake was used to trick an employee at a Hong Kong-based firm into transferring about $25 million to a fraudster after the employee thought he was conferring with the CFO of the company on video call.

Lawsuits over injury stemming from AI use is also a major category, Lewin-Smith said. OpenAI is facing multiple lawsuits alleging that ChatGPT users were encouraged by the system to take their own lives. Many insurers are not even aware that they could be exposed to those types of risks, the Testudo CEO said.

The market for generative AI coverage is a "fairly unique niche," Lewin-Smith said. Testudo recently launched a new insurance product intended to help protect entities against the fallout that may occur when a company's GenAI system "causes harm."

Traditional underwriters are becoming concerned about their exposure to generative AI as they pay out claims "silently covering a risk that they don't really understand," said Lewin-Smith. "We're seeing more startups come in to try and fill gaps and all have a different approach."

Accelerating innovation

Testudo was part of a 10-week accelerator program at Lloyd's, and Rosie Denée, head of innovation and commercial education at the market, told S&P Global Market Intelligence that its GenAI liability coverage is a "super-interesting product."

"Their pitches to investors and market experts were really well-received and we're very proud that they have successfully launched new coverage for liability risks created by Gen AI," she said.

"We can turn around a price and quote within an hour based on high-level factors about your base model provider, your business application and your business more broadly," Lewin-Smith said.

Testudo has proprietary "active liability monitoring" technology that scans US litigation data sets and then analyzes where the liability risk is coming from. The company then works to lower that risk, similar to how home insurers might give customers information to harden their homes against wildfires.

Companies should check with their insurers to see if they are covered under their current policies for their use of AI, a question he suspects will result in a "pretty underwhelming" answer, Lewin-Smith said.

Using AI is "critical to long-term competitiveness" for many companies but they need to make sure they have proper "risk-transfer mechanisms in place" to deal with potential liabilities. "That's what we're offering," said Lewin-Smith.