Insurers Lack Verification Systems as AI-Generated Fraud Rises

Insurance companies are rapidly automating claims and underwriting decisions while lacking formal systems to verify whether photos, documents, and voice recordings have been manipulated by artificial intelligence, according to new research examining the industry's preparedness for synthetic fraud.

A study commissioned by Clearspeed and authored by insurance innovation strategist Sabine VanderLinden analyzed 76 public filings from 49 insurers and reinsurers, along with 31 industry studies and interviews with 16 claims and underwriting executives in the U.S. and U.K. The findings reveal a significant gap between carriers' AI adoption and their fraud detection capabilities.

Public Disclosures Ignore Synthetic Evidence

Researchers searched annual reports, 10-K filings, proxy statements, and statutory returns for references to AI-generated evidence, manipulated media, and synthetic identities. None of the 76 documents mentioned synthetic media, synthetic identity, or voice cloning. Only six of the 49 companies referenced deepfakes, and exclusively in cybersecurity contexts rather than claims or underwriting verification.

The disconnect extends to reinsurers. Among the five largest global reinsurers examined, none mentioned deepfakes, synthetic media, or AI-generated evidence in their most recent annual reporting.

Yet the threat is tangible. Separate industry research from March 2026 found that 98% of 300 U.S. insurance claims professionals surveyed agreed AI editing tools are fueling digital media fraud. Despite this consensus, only 32% expressed confidence in their ability to identify a deepfake.

The Trust Deficit's Business Impact

Fraud represents approximately 10% of property and casualty losses, according to the Coalition Against Insurance Fraud. The organization estimates the broader trust deficit costs the U.S. insurance system at least $308.6 billion annually.

The verification challenge affects legitimate customers as well. Ian Thompson, former group chief claims officer at Zurich Insurance, told researchers that more than 90% of claimants are honest. Fraud detection efforts aimed at the remaining minority can leave legitimate policyholders feeling distrusted, creating friction in the customer experience.

Researchers noted that insurers face a fundamental challenge: determining which interactions warrant expedited processing, which require heightened scrutiny, and which demand human judgment.

Proposed Framework for Verification

The report, titled "The Speed of Trust: Building the Trust Intelligence Layer for Insurance in the Age of Agentic AI," proposes that insurers establish what VanderLinden calls a Trust Intelligence Layer—a continuous, auditable risk indicator that operates throughout the policyholder journey.

This framework would inform decisions rather than automate them, produce audit trails instead of automated denials, and function without requiring demographic or historical data about the individual being assessed. The goal is to clear legitimate interactions quickly while directing human review toward exceptions.

Why it matters

As insurers deploy AI agents to handle customer interactions and process claims, they're creating systems that can be exploited by increasingly sophisticated synthetic fraud. Without verification infrastructure that matches the sophistication of their automation, carriers risk both financial losses from undetected fraud and customer attrition from overly aggressive screening. The absence of any mention of synthetic media threats in regulatory filings suggests the industry may be underestimating a risk that its own claims professionals recognize as significant.

The research was first reported by ProgramBusiness.com and independently authored by Sabine VanderLinden of Alchemy Crew Ventures.