Insurtech—the use of technology to change insurance administration, underwriting, sales and claims—is already part of health and life insurance. Insurers report using or exploring artificial intelligence (AI) and machine learning (ML) for tasks such as prior authorization, fraud detection and life underwriting. But growing use is not proof of better results for policyholders: available evidence does not establish that these tools have broadly lowered premiums, reduced coverage denials or improved health outcomes.
How widely are insurers using AI and other digital tools?
Recent regulator surveys indicate adoption, but their findings describe different groups and should not be combined into a single global estimate.
| Survey finding | Who it covers | What it tells you |
|---|---|---|
| 84% reported using AI/ML in some capacity; nearly 92% reported AI/ML governance principles modeled on NAIC principles. | 93 health insurance companies surveyed by 16 U.S. states from November 2024 to January 2025. The companies were selected under stated premium or market-share criteria; the survey was not a census of all insurers. Source: National Association of Insurance Commissioners (NAIC), 2025. | AI/ML use and reported governance are common among respondents, not necessarily among every U.S. health insurer. Governance policies do not by themselves prove a system is accurate, fair or harmless. |
| 50% of non-life respondents and 24% of life respondents reported AI use. Nearly 80% of respondents used BigTech firms for cloud storage. | Respondents to EIOPA’s European digitalisation monitoring, reported in 2024. Source: European Insurance and Occupational Pensions Authority (EIOPA), 2024. | These are findings from a European respondent group, not all insurers. EIOPA’s non-life category is not the same as the NAIC survey’s health-only group. |
The NAIC respondents also reported testing for model drift and bias, checking accuracy and data quality, conducting equity and compliance audits, and using human oversight. These are self-reported practices, not independent evaluations of particular systems. EIOPA described IoT, blockchain and parametric insurance as technologies used by only a small number of insurers in its survey.
Where is technology changing health insurance?
Health insurers responding to the NAIC survey reported using or exploring AI/ML in several operational and coverage-related areas. The survey does not mean every insurer uses every application.
- Plan administration and service: online sales, quoting and shopping can automate parts of enrollment and customer interactions.
- Care and utilization management: tools may support disease-management programs and review requests, including prior authorization.
- Fraud detection: systems can flag suspected fraud involving claims or providers for further examination.
These applications have different stakes. Automating a routine administrative task is not the same as using a model to influence whether a service is covered or a patient can access care. For a consumer, the important questions are what the system is allowed to decide, whether a person reviews consequential decisions, and how an affected member can ask for an explanation or challenge a result.
How can digital records affect life insurance underwriting?
Life insurers assess application information, including relevant health history. Electronic health records (EHRs) can contain complex, inconsistently formatted information, so one possible use of language-processing technology is to organize records for underwriters rather than make a final decision on its own.
A provider-described EHR summarization product
In an April 23, 2024 announcement, Munich Re Life US and Clareto described an Automated EHR Summarizer that extracts and normalizes EHR data, highlights information relevant to underwriters and offers triage guidance. The announcement said it produces both a human-readable HTML report and structured digital data for rules, models and analytics. It described possible uses in accelerated underwriting, post-issue audit and light-touch underwriting.
The companies said the product was available on April 23, 2024. That announcement does not establish its current availability, independent accuracy, effect on application times or effect on consumers’ decisions. A product description is evidence of what its providers say the tool can do, not proof of customer benefit. Source: Munich Re, April 23, 2024.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →What could these tools mean for policyholders?
Digitized workflows may help insurers organize information, route cases or flag records for review. In principle, that could change how quickly or consistently some tasks are handled. However, the regulator surveys and provider announcement cited here do not establish that insurtech has broadly made underwriting more accurate, lowered premiums, reduced denials or improved health outcomes. Those are consumer outcomes that require evidence beyond adoption figures or a product’s stated capabilities.
AI can also create risks when data is incomplete, inaccurate or used in ways that are difficult to explain. A model can reflect bias in its data or produce results that shift over time. When a tool affects coverage, pricing or underwriting, the practical issue is not simply whether AI is involved; it is what data and decision process are used, how they are checked, and what recourse a consumer has.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What safeguards and rules apply?
Insurance regulation depends on jurisdiction, and the cited materials do not establish one universal AI rule.
European Union
EIOPA’s August 2025 opinion says AI systems used for risk assessment and pricing in life and health insurance are high-risk under the EU AI Act. The opinion addresses national supervisors and clarifies how existing insurance-sector legislation applies in the context of AI; EIOPA says it does not create new requirements or change the scope of existing legislation. Its supervisory considerations include data governance, record-keeping, fairness, cybersecurity, explainability and human oversight. Source: EIOPA, 2025.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Wisconsin
A Wisconsin Office of the Commissioner of Insurance bulletin dated March 18, 2025, expects insurers it regulates to develop and maintain a written AI systems program for AI supporting decisions related to regulated insurance practices. It calls for controls proportionate to the decision, potential consumer harm, human involvement, explainability and use of third-party data or systems. The guidance also discusses data quality and lineage, bias analysis, validation, monitoring, documentation and consumer notice. This is Wisconsin guidance relying on Wisconsin law, not a nationwide mandate. Source: Wisconsin OCI, March 18, 2025.
United States survey findings
The NAIC survey provides a snapshot of reported practices among participating companies, not a guarantee that an insurer’s system meets a particular standard or that a consumer is protected from every error. Pennsylvania Insurance Commissioner Michael Humphreys, then chair of the NAIC Big Data and Artificial Intelligence (H) Working Group, called survey completion “a key milestone” and said regulators’ work was not done. Source: NAIC, May 20, 2025.
What should you ask about an insurer’s technology?
If a digital tool is involved in a decision about your health coverage or life insurance, these questions can help clarify what it does and how the insurer handles its risks:
- Purpose: Is the tool used for administration, prior authorization, fraud detection, life underwriting or claims—and does it assist a person or determine an outcome?
- Data: What information does it use, how was that information obtained, and what privacy, consent, quality and security controls apply?
- Review and recourse: Can a qualified person review or escalate the result? How can you ask for an explanation, correct inaccurate information or challenge a decision?
- Testing: How does the insurer check accuracy, bias, missing data and changes in model performance after deployment?
- Outside providers: Does a third-party vendor supply the tool or data, and how does the insurer assess and audit that provider?
- Evidence: What independent evidence shows an effect on access, customer experience, accuracy, costs or outcomes?
These questions distinguish a technology’s advertised capabilities from the controls and results that matter to a policyholder. The cited regulator materials and product announcement do not provide a head-to-head vendor comparison or an independent cross-market assessment of consumer outcomes.
Recommended Free Tools
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
