Insurers are automating claims and underwriting faster than their verification systems are developing, according to new research commissioned by Clearspeed. The study found a widening gap between AI adoption and insurers’ ability to validate information used in automated decisions. Researchers examined how insurers handle evidence as AI takes on more decisions, customer interactions and workflow handoffs.
The research covered 76 public filings from 49 insurers and reinsurers alongside 31 insurance studies. Researchers also conducted 16 interviews with claims and underwriting executives at insurance companies in the US and UK.
Insurers are increasingly automating evidence reviews and customer interactions as generative AI makes manipulated information easier to produce. Photos and documents now require greater scrutiny, and synthetic voices or identities create another source of fraud exposure, Beinsure noted.
Researchers searched annual reports, 10-K filings, proxy statements and statutory returns for references to AI-generated or manipulated evidence. None of the 76 filings mentioned synthetic media, synthetic identities or voice cloning.
Six of the 49 insurers discussed deepfakes, though those references appeared only within cybersecurity disclosures. None addressed deepfakes as evidence entering claims or underwriting workflows.
The research separately reviewed recent reporting from five of the world’s 10 largest reinsurers. None mentioned deepfakes, synthetic media or AI-generated evidence in their latest annual disclosures.
Claims professionals already recognize the problem more clearly at an operational level. Separate industry research published in March 2026 surveyed 300 US insurance claims professionals and found 98% agreed that AI editing tools are increasing digital media fraud.
Their confidence in detecting manipulated material was much lower. Only 32% said they were highly confident in their ability to identify a deepfake.
The figures expose a practical issue for claims departments as more evidence arrives digitally. AI reduces the cost and technical difficulty of producing convincing images, documents or audio, while insurers are simultaneously reducing manual review across routine claims.
Insurance fraud already represents a substantial expense. The Coalition Against Insurance Fraud estimates fraud accounts for about 10% of property and casualty losses, while the wider trust deficit costs the US insurance system at least $308.6 bn annually.
Verification processes create costs for legitimate policyholders as well. Insurers must decide which interactions deserve immediate processing and which require more scrutiny or human judgment.
Researchers that more than 90% of customers filing claims are honest. Fraud controls aimed at the remaining minority still risk making legitimate customers feel distrusted.
That tension becomes more pronounced as automation expands. Excessive verification slows genuine claims, while weak verification gives manipulated evidence a clearer path through automated workflows.
Insurers are automating claims and underwriting faster than their verification systems are developing, according to new research commissioned by Clearspeed. The study found a widening gap between AI adoption and insurers’ ability to validate information used in automated decisions. Researchers examined how insurers handle evidence as AI takes on more decisions, customer interactions and workflow handoffs.
Insurers are increasingly automating evidence reviews and customer interactions as generative AI makes manipulated information easier to produce. Photos and documents now require greater scrutiny, and synthetic voices or identities create another source of fraud exposure, Beinsure noted.Researchers searched annual reports, 10-K filings, proxy statements and statutory returns for references to AI-generated or manipulated evidence. None of the 76 filings mentioned synthetic media, synthetic identities or voice cloning.
The research separately reviewed recent reporting from five of the world’s 10 largest reinsurers. None mentioned deepfakes, synthetic media or AI-generated evidence in their latest annual disclosures.Claims professionals already recognize the problem more clearly at an operational level. Separate industry research published in March 2026 surveyed 300 US insurance claims professionals and found 98% agreed that AI editing tools are increasing digital media fraud.
Their confidence in detecting manipulated material was much lower. Only 32% said they were highly confident in their ability to identify a deepfake.
Insurance fraud already represents a substantial expense. The Coalition Against Insurance Fraud estimates fraud accounts for about 10% of property and casualty losses, while the wider trust deficit costs the US insurance system at least $308.6 bn annually.Verification processes create costs for legitimate policyholders as well. Insurers must decide which interactions deserve immediate processing and which require more scrutiny or human judgment.
That tension becomes more pronounced as automation expands. Excessive verification slows genuine claims, while weak verification gives manipulated evidence a clearer path through automated workflows.