The insurance industry has always adapted to new tools to improve underwriting. From boots-on-the-roof inspections to high-resolution satellite imagery, the goal has remained to evaluate risk accurately and price it appropriately. But as I noted in I Was Droned: California Policyholders Are Being Monitored Regarding Their Loss Risk Exposure and Loss Mitigation Attempts, something has fundamentally changed. The scale, frequency, and lack of transparency of today’s insurance surveillance tools are unlike anything we have seen before.
We are no longer talking about a field inspector taking a few photographs during underwriting. We are now in an era where insurers can continuously monitor properties using satellites, drones, and AI-driven analytics. At what point does efficient underwriting cross into unfair or unregulated surveillance?
To their credit, insurance regulators are beginning to grapple with that very issue. The Colorado Division of Insurance’s newly issued Bulletin B-5.57 is one of the clearest statements yet that while aerial imagery is useful, it is not infallible and should not be treated as determinative. The bulletin recognizes that what many policyholders have already experienced firsthand, images can be outdated, misleading, or simply wrong. A shadow becomes “roof damage.” A tarp becomes “neglect.” A temporary condition becomes grounds for nonrenewal.
The Colorado bulletin draws a firm line that aerial imagery alone is not enough. It cautions insurers against relying on imagery as the sole basis for adverse underwriting or claims decisions and requires additional verification when conclusions are uncertain. That is a significant and welcome development.
For me, transparency is key, and the most important part of the bulletin is the principle that goes to the heart of fairness:
Policyholders and applicants impacted by an adverse action supported by aerial imagery should be given a meaningful opportunity to dispute the accuracy of aerial imagery, correct errors, provide updated information, and submit proof of completed repairs or remediation. Insurers should review information submitted by policyholders and applicants disputing the accuracy of aerial imagery and consider whether further evaluation or reconsideration is warranted.
Transparency and the proper use of technology is where the conversation needs to be. The issue is not whether insurers should use technology. Of course they should. As I discussed in The Future of Homeowners and Property Insurance: Navigating AI, Surveillance, and Regulatory Challenges, the insurance industry is rapidly moving toward AI-enhanced decision-making. That trend is not going away. The real question is whether the guardrails protecting policyholders are keeping up.
Other insurance regulators are sounding similar alarms. Michigan regulators have issued guidance emphasizing transparency and cautioning against overreliance on third-party data models that consumers cannot see or challenge. The NAIC has been working through model guidance focused on algorithmic accountability, data governance, and consumer transparency. NCOIL has also explored legislative responses addressing the use of aerial imagery and the need for fairness in underwriting practices.
What distinguishes the Colorado bulletin is its clear articulation of the technology’s limits. It acknowledges that aerial imagery has “surface-only limitations” and may require enlargement, extrapolation, or inference that can lead to error. In other words, it recognizes that what looks obvious from 500 miles above the earth may be anything but accurate when you are standing on the roof.
This is not an academic concern. Across the country, policyholders are receiving nonrenewal notices or coverage restrictions based on imagery they have never seen and cannot easily challenge. That lack of transparency undermines trust in the insurance relationship.
Do insurance industry leaders understand that widespread distrust in the insurance product is bad for the long-term health of the insurance industry? The industry’s defenders often argue that this technology is necessary because consumers misrepresent risks or are unaware of property conditions. There is some truth to that. But that reality does not justify a system in which decisions are made based on hidden, unverified, and effectively unchallengeable data. Why the secrecy?
The Colorado bulletin moves the conversation in the right direction by recognizing that aerial imagery is a tool and not the final word. If technology is used to make decisions about coverage, pricing, or claims, policyholders must have a fair opportunity to see, understand, and respond to that information. What is the objection to that?
Thought For The Day
“The great aim of the law is not to punish but to prevent injustice.”
— Oliver Wendell Holmes Jr.
