Physical AI Is Forcing Insurers to Start From Scratch

Physical AI Is Forcing Insurers to Start From Scratch

Physical artificial intelligence (AI) doesn’t follow a script. Unlike traditional machines that execute fixed commands, physical AI systems use sensors, software and AI models to decide how to act as conditions change. That flexibility is also what breaks the liability framework traditional policies was built to handle.

When something goes wrong, the fault rarely belongs to one company. A robot maker may have built the hardware. A separate firm may have supplied the AI model. The operator may have set failed to enforce the operating rules. Contracts divide those duties on paper, but a single incident can still trigger disputes over which failure caused the loss.

Existing policies leave gaps on both ends of a robot incident. A standard business insurance policy may not cover the loss at all if it includes an AI exclusion or if the loss doesn’t fit its definition of an accident. If it does respond, it likely covers only physical damage, not the production losses that follow when a robot failure shuts down a line.

Insurers are choosing to exclude AI from their policy coverage. Berkshire Hathaway, Chubb and Travelers sought state regulatory approval to exclude AI-related damages from general liability policies and state regulators approved more than 80% of those requests, PYMNTS reported. Some carriers went further, establishing absolute exclusions across multiple policy lines.

New Robot-Specific Policies Fill Gaps Standard Coverage Misses

Axis Insurance built a program specifically for companies that make and deploy autonomous robots. It covers bodily injury and property damage from AI navigation or perception failures, physical damage caused by a cyberattack that takes over a robot’s controls, and production losses from a software update or sensor failure that takes a robot offline, even when no physical damage has occurred. It also covers cases where a defect in a third-party sensor or component triggers a failure, but the claim lands on the company that integrated it rather than the part maker.

Last January, Relm Insurance launched a product called PONTAAI that covers bodily injury and property damage among other liabilities.

Advertisement: Scroll to Continue

Neither product covers the full chain of liability. A single robot incident can trigger claims against the hardware maker, the software developer and the company operating the machine.

Insurers Lack Claims History to Price Autonomous Robot Risk

Insurers usually price commercial risk using past claims, equipment records and operating controls. Physical AI gives them less history and more variables. The same robot can carry different risk at two warehouses. Floor layouts differ. Workers interact with the system in different ways. Software versions and operating limits change after the policy is written.

The pattern matches what the industry saw with cyber insurance, when adoption outpaced the understanding of how risk translated into insured loss, Lucy Pilko, CEO Americas at AXA XL, wrote in a blog post.

The accumulation problem runs deeper still. Gallagher Re found that as robots and autonomous vehicles increasingly run on AI, product liability coverage may respond to injury and property damage, but only in jurisdictions that treat AI software as a product. Financial losses from the same failure get no coverage at all.

That risk has no geographic limit. A hurricane is bounded by landfall. A factory fire is bounded to one site. A flaw in one widely used AI model can spread instantly to every business running it, in every industry and country at once, according to Gallagher Re.