AI Mistakes: Will Insurance Cover the Cost?
Facing mounting liabilities, major insurance firms are rethinking how they cover AI-related mistakes. How that could upend firms’ use of the tech.
If mistakes made by a company employee caused damages, the firm’s insurance would likely cover most—if not all—of the costs. But who covers the damages if it’s the organization’s AI that makes the mistake?
As if the world of AI hasn’t been confusing enough, now comes a new thorny issue: Big insurers, which mostly have no qualms about covering damages from fires, product defects, and even fraud, are now thinking twice about being backstops for AI’s mistakes. AI-related lawsuits increased by nearly 1,000% from 2021 to 2025, and by 137% from 2024 to 2025 alone, according to insurance broker Gallagher. Several major US business insurers now exclude liabilities tied to chatbots, agents, and other AI tools from their traditional business coverage. Numerous other carriers have filed with state regulators asking for approval to exclude AI risk. “A business may think, ‘I’m covered,’ but that may not be true,” says Liam Lawrence, a Korn Ferry senior client partner and the firm’s practice leader for Insurance in North America.
Though it doesn’t get much notice, commercial liability for the insurance industry is no small matter. Some estimates put it at over $175 billion last year—more than the liability for all natural disasters combined. Only a couple of years ago, AI seemed mostly to be a technology-company tool, but today almost every business is using it for something, be it hiring, marketing, customer service, analytics, or even decision-making. While right now many are fixated on whether a near-future AI model could cause a civilization-ending disaster, there are plenty of smaller risks associated with the models already in wide use. “The potential liability is much broader,” Lawrence says. It’s likely that the issue will only grow in importance as more firms increase their use of AI. Yet it isn’t on the radar for many AI-using firms, says Shanda Mints, Korn Ferry’s vice president of AI strategy and transformation, and perhaps it should be. “AI can magnify the impact of a wrong decision or direction,” she says.
No technology is perfect, but even mistakes that seem minor can be financially explosive if AI is involved. For instance, a company’s AI could unintentionally give customers bad advice or hack another organization’s systems. Indeed, the ubiquity of AI is the reason so many insurers are struggling with handling it. Thousands of seemingly independent risks, from an assembly line breaking down to employees getting bad advice on a 401(k)—and everything in between—could all be tied to a single AI model. “AI concentration risk is the strategic AI insurance risk,” says Christopher Orr, a Korn Ferry senior client partner in the firm’s Global Insurance practice.
Compounding the issue is that a single AI-related problem could be characterized in so many different ways: as a security incident, a professional error, a defective product, a discriminatory decision, a governance failure, fraud, or some combination of any or all of these. “The insurance question is not only whether AI creates bigger losses, it’s whether AI creates connections across their portfolio they didn’t realize were there,” Orr says. The implication is that a single organization-wide AI policy likely won’t work.
The insurance industry has periodically had to develop coverage for new or worsening kinds of events. After the September 11 attacks, insurers spent several years rewriting policies to address terrorism risk. As the frequency of catastrophic natural disasters has increased, carriers have had to modify policies for hurricanes and wildfires. AI-related liability represents the latest challenge the industry has faced, Lawrence says: “The only difference is how quickly it’s happening.”
Companies that feel too exposed may have to reduce their AI use until coverage becomes available, experts say. Smaller insurers are already rushing in to provide AI-specific policies, while bigger carriers are starting to offer riders to existing ones. Either way, AI governance needs to become a board-level issue, not simply a technology function, Orr says.
Learn more about Korn Ferry’s AI in the Workplace capabilities.
