Aerial and satellite imagery in underwriting
Using images taken from aircraft, drones or satellites, analyzed by models, to assess a property's condition and surroundings without a visit, roof, vegetation, pool, and to estimate damage after a catastrophe.
High-resolution aerial and satellite images, now refreshed frequently, make it possible to observe an insured property without going there. Computer vision models extract features useful to the underwriter: roof condition and material, presence of solar panels, pools or undeclared extensions, actual building footprint, distance to combustible vegetation, exposure to water. In underwriting, this information checks or supplements statements and reveals changes during the contract. In claims, it makes it possible to estimate within days, after a hurricane, wildfire or flood, which buildings are destroyed or damaged, and to prioritize interventions and payments. The use raises questions of proportionality and fairness: a policyholder whose contract is canceled or surcharged on the strength of an image must be able to challenge it, and several US regulators have reacted to non-renewals based on misread or outdated images. The method is powerful for physical risk, but it sees neither building interiors nor the quality of maintenance, and must remain one element of assessment among others.
After the Los Angeles area fires in January 2025, insurers analyzed aerial images taken in the following days to classify thousands of insured addresses as destroyed, damaged or intact, and paid advances on destroyed properties before any adjuster visit, in areas that remained inaccessible for a long time.
Follow the research
One letter a week. First, the week's underwriting note, when one is published: a dated, argued decision, signed by Paul Bernardski. Then what the observatory published on cyber risk, political risk and insurance. No account, no tracking.
No spam, one-click unsubscribe, no data ever sold.