Newsletters
Risk Central
Digital EditionInsurers Broaden A.I. Adoption as Benefits Build Over the Long Term
North American insurers are expanding their use of artificial intelligence across underwriting, pricing, claims, distribution and operations, with approximately 34% of respondents implementing A.I. broadly throughout their organizations, according to a Moody’s Ratings report.
Moody’s surveyed 58 insurance companies spanning life, health, property and casualty, reinsurers, and brokers. The ratings agency found that roughly two-thirds of respondents had multiple production use cases across core operations, while the remainder reported more limited production use in selected functional areas.
Moody’s said it expects the benefits of A.I. to emerge over time, enhancing efficiency and revenue growth, though outcomes will vary significantly across companies and insurers slow to adopt the technology risk greater competitive pressure and earnings volatility.
A.I. Use Spans the Insurance Value Chain
Moody’s found that A.I. is being used across nearly every stage of the insurance business, with generative A.I. tools such as large language models, copilots and chatbots cited by 98% of respondents, followed by traditional machine learning and predictive models at 79%.
Among life insurers, operations showed production use at 50%, while customer service and distribution had the highest level of production use among the functions surveyed, at 66%, followed by underwriting and risk selection. P&C insurers and reinsurers reported the heaviest production use in operations at 75%, with claims and loss adjustment, fraud detection, and underwriting and pricing also showing substantial deployment.
Respondents cited reduced processing and cycle times and headcount avoidance or productivity gains as the most common benefits, at 74% and 76% of respondents, respectively, while P&C insurers also pointed to reduced loss severity, lower loss adjustment expense and improved combined ratios. Moody’s said benefits have largely been tracked for individual use cases, with 41% of respondents quantifying and tracking results, 34% quantifying but not consistently tracking, and 24% describing benefits as primarily qualitative.
Insurers are also applying A.I. to investment portfolio management, with 43% of respondents deploying tools directly and 22% doing so indirectly through external managers; faster research and decision-making was the most cited benefit, at 55%.
Spending Levels Could Widen the Gap Between Large and Small Insurers
A.I. spending represents roughly 2% to 10% of total information technology budgets on average, Moody’s found, with 13% of respondents reporting that A.I. spending exceeds 10% of their overall IT budget. P&C insurers reported notably higher spending levels than life insurers: 19% of P&C respondents characterized their spending as significant, compared with 7% of life insurers, while 44% of P&C insurers reported meaningful spend versus just 3% of life insurers.
Moody’s said larger insurers are generally better positioned to absorb the investment required because of greater technology resources and broader data pools, while smaller companies may face greater implementation constraints. The rating agency said continued investment will be important for insurers to maintain competitiveness, and it views the overall credit impact as neutral to positive, with outcomes for individual companies depending on execution quality, data capability and governance rather than business model displacement.
About 55% of respondents said they selectively measure returns on A.I. investments and expect to evaluate benefits within three years, though many insurers have not formally defined a time frame for measuring returns, which Moody’s attributed in part to companies still being in an experimental phase.
Governance Gaps Could Expose Insurers to New Risks
Survey respondents identified skills and talent constraints as the leading obstacle to realizing A.I. benefits, cited by 81%, followed by data quality and availability at 71% and integration with legacy systems at 62%. Regulatory uncertainty was cited by 47% of respondents, which Moody’s attributed to multiple and sometimes divergent regulatory frameworks governing A.I. use across jurisdictions.
Despite these challenges, 88% of respondents said their A.I.-specific governance principles were aligned with regulatory guidance, while only 5% said they were still developing governance principles or had none in place. Moody’s said embedding A.I. into core processes increases operational, regulatory and litigation risk, deepens dependence on key vendors, and heightens tail risks tied to data privacy, quality and security, along with greater exposure to cyberattacks, fraud and model failures. &
