Incorporating AI Risks into Underwriting Decisions by 2028

ScienceSoft, an IT consulting and software development firm, has published research forecasting a significant shift in how insurers will incorporate AI risks into underwriting decisions by 2028.

According to ScienceSoft's report, by 2028, up to 80% of new insurance policies and renewals in areas such as errors and omissions, directors and officers, employment practices liability, and cyber insurance will integrate AI risks into their underwriting processes. This not only reflects a growing awareness of AI-related risks but also signals a profound change in underwriting strategies across several insurance lines. The transformation highlights the urgency for insurers to reassess how they perceive and manage emerging risks associated with rapid technological advancements.

AI-Specific Products vs. Policy Adaptations

Despite these forthcoming changes, many medium-sized insurers in the U.S. are likely to tackle AI-related risks using existing policy frameworks rather than developing distinct AI-specific products. The market for AI-specific insurance is projected to grow substantially from $40 million in 2024 to a staggering $4.8 billion by 2032. However, it will remain a minor component of the broader commercial property and casualty insurance market, representing just 0.34% of the total premiums.

This trend underscores an industry-wide inclination towards adapting traditional policies with new exclusions and policy language to address AI risks, rather than creating entirely new product lines. The study corroborates findings from Insurance Business, which identified similar trends in insurance filings and policy revisions nationwide.

Emerging AI Exclusions and Adaptations

Recent developments further exemplify this trend. In 2025, the Insurance Services Office introduced standard generative AI exclusion forms, which numerous insurers adopted by 2026 for lines like commercial general liability and errors and omissions. Leading reinsurers, including W.R. Berkley, Chubb, and AIG, have either aligned with standard exclusions or developed proprietary versions, indicating wide industry acceptance.

However, uptake varies across market segments. Employment practices liability insurers show less frequent AI exclusions, and many directors and officers insurers have not yet broadly implemented such exclusions. According to Gallagher's 2026 AI Adoption and Risk survey, while 20% of insurance professionals report clients experiencing AI-related losses, less than half have formal AI risk management frameworks.

AI Liability Products & Market Realities

While exclusion trends dominate, some insurers like HSB are moving towards offering affirmative AI liability products. This development supports ScienceSoft's projection of a maturing, albeit niche, market for AI-specific insurance solutions. ScienceSoft's insight into AI incident frequency further aligns with reports from the AI Incident Database, illustrating a surge in incidents prompting detailed underwriting policies.

Key Takeaways for Insurance Professionals

For insurance carriers and brokers, the ongoing scrutiny of AI governance and control measures is of substantial importance. While ScienceSoft offers valuable projections, industry stakeholders must approach these insights judiciously, considering independent analyses and diverse market perspectives. As the insurance landscape evolves, staying informed on AI advancements and their implications for underwriting and risk management remains crucial.