The rapid integration of artificial intelligence into the core operations of global enterprises has reached a critical juncture for the insurance industry. As businesses across all sectors adopt generative AI and machine-learning algorithms to streamline workflows, insurers are moving with unprecedented speed to limit their exposure. Recent trends indicate that the commercial insurance market is shifting away from "silent" coverage of AI-related risks toward explicit exclusions, a move driven by a staggering increase in litigation and the introduction of new standardized policy language.
According to industry analysts and legal experts, the insurance industry is currently in the midst of a transformative reaction to the AI explosion. This shift is characterized by a surge in filings with state insurance regulators as carriers seek to adopt standardized exclusions developed by the Insurance Services Office (ISO), a subsidiary of Verisk. These exclusions are designed to provide underwriters with the precision necessary to navigate a landscape where the potential for liability is both vast and largely untested in the courts.
The Evolution of AI Risk and the Shift Toward Explicit Exclusions
For several years, artificial intelligence existed in a state of "silent coverage" within many commercial general liability (CGL), professional liability, and cyber insurance policies. Because AI was not specifically mentioned, many legal experts argued that traditional policy language could be interpreted to cover losses resulting from algorithmic errors, data breaches, or intellectual property disputes involving AI. However, as the technology transitioned from a niche tool to a ubiquitous business necessity, the potential for catastrophic, aggregate losses became a primary concern for actuaries and risk managers.
The timeline of this shift has been remarkably compressed. While the insurance industry typically moves at a deliberate pace, the rise of generative AI in late 2022 served as a catalyst for immediate action. By 2023, early adopters in the insurance space began drafting bespoke exclusions. By 2024, the demand for standardized language led to the development of the three primary ISO endorsements that are now being filed across various jurisdictions.
Alana McMullin, a partner at Lathrop GPM who specializes in complex insurance disputes and toxic tort litigation, observes that the industry is currently in the middle of a major defensive pivot. "There’s been a major shift in the insurance industry’s treatment of AI-related risks," McMullin noted. "Insurers are moving very quickly to limit this exposure. These ISO exclusions were the spark of this AI exclusion boom that’s happened now, and we’re in the middle of what I’d call an industry-wide reaction to the explosion of AI."
Analyzing the ISO Endorsements and Underwriting Flexibility
The ISO exclusions have become the focal point for carriers looking to stabilize their books of business. Joe Lam, vice president of liability at Verisk and a key contributor to the development of these endorsements, emphasizes that these tools are essential for maintaining a healthy marketplace. Without the ability to explicitly exclude or define AI risks, underwriters might choose to withdraw from certain high-risk sectors entirely rather than risk unforeseen exposure.
The standardized endorsements provide carriers with a menu of options. Some allow for a total exclusion of AI-related liabilities, while others might offer limited exceptions or allow for the "buy-back" of certain coverages through additional premiums. This flexibility is intended to prevent a market freeze. As Lam explains, insurance is an exchange of risk for premium, and that exchange requires a clear understanding of what is being covered. "Exclusions are very essential in the marketplace," Lam said, noting that they allow carriers to remain in the market by providing a mechanism to manage emerging issues that were never contemplated when original policy forms were drafted.
The move toward these exclusions is not merely a defensive posture; it is also a prelude to more sophisticated underwriting. By first excluding the broad, undefined risk of AI, insurers can then begin to offer specific, priced-out coverage for policyholders who can demonstrate robust AI governance and risk management protocols.
The Litigation Landscape: A 978% Surge in AI Lawsuits
The motivation for this rapid adoption of exclusions is grounded in hard data. A recent study by Gallagher, utilizing data from the AI underwriter Testudo Global, Inc., revealed a seismic shift in the legal environment. Between 2021 and 2025, AI-related lawsuits saw a staggering 978% increase. Even more concerning for insurers was the 137% increase in litigation observed between 2024 and 2025 alone.
This surge in legal action is not limited to a single area of law. Instead, AI-related claims are spanning a wide spectrum of liability:
- Patent and Intellectual Property Infringement: Accounting for 11.9% of cases, these claims challenge the very foundation of how AI systems are built. They involve disputes over the functionality of algorithms and the proprietary nature of the code.
- Copyright Infringement: Comprising 11.2% of cases, this category focuses on the data used to train large language models. Content creators, authors, and media companies are increasingly suing AI developers for using copyrighted material without authorization.
- Personal Injury and Privacy: Making up 10.2% of cases, these claims involve privacy violations, the misuse of personal data, and what legal scholars are beginning to call "digital dignity" violations.
- Consumer Protection and Discrimination: Lawsuits are increasingly targeting the use of AI in hiring, lending, and housing, alleging that biased algorithms lead to discriminatory outcomes.
McMullin points out that while the technology is novel, the underlying legal theories—such as disclosure issues, violations of governance policies, and misrepresentations to shareholders—are familiar. These are particularly dangerous for Directors and Officers (D&O) and Errors and Omissions (E&O) lines of coverage.
Case Study: Berkley’s Absolute AI Exclusion
One of the most prominent examples of the industry’s aggressive stance is the "absolute AI exclusion" introduced by W.R. Berkley. This exclusion, designed for use in D&O, E&O, and fiduciary liability products, represents one of the most comprehensive attempts to insulate a carrier from AI risk.
The Berkley exclusion reportedly goes beyond simply excluding the use of AI. It encompasses the development of AI, the generation or dissemination of content using AI, and even the failure of a policyholder to identify content generated by a third party’s AI. Furthermore, it excludes coverage for an insured’s internal policies and procedures regarding AI, as well as any breach of duty or legal obligation related to the technology.
This "absolute" approach serves as a benchmark for the industry. While not every carrier will adopt such a broad exclusion, the existence of such language in the market sets a precedent for how far insurers are willing to go to protect their balance sheets from the unpredictable "tail risk" associated with autonomous systems.
Market Dynamics: Exclusion vs. Innovation
Despite the trend toward exclusions, the insurance market remains a competitive environment. Some experts believe that market pressure may temper the widespread adoption of the most aggressive exclusions. If a carrier implements an exclusion that is too broad, it may find its policies unattractive to modern businesses that rely on AI for their daily operations.
"Aggressive exclusions may make their policies less attractive, especially given the pervasive nature of AI in today’s business environment," McMullin noted. She suggests that some insurers may take the opposite approach: instead of excluding the risk, they may choose to embrace the uncertainty by underwriting it at a premium. This creates a bifurcated market where some businesses are forced to accept exclusions, while others pay significantly more for specialized "AI-inclusive" policies.
The ultimate outcome will likely be decided by policyholders during the renewal process. Large corporate clients with significant leverage may be able to negotiate "carve-outs" or limited exceptions to AI exclusions, provided they can prove their AI usage is governed by strict ethical and technical standards.
Analysis of Implications and Future Outlook
The industry-wide move toward AI exclusions marks the end of the era of "silent AI" and the beginning of a more rigorous, albeit more expensive, risk management landscape. For businesses, the implications are clear: they can no longer assume that their existing liability towers will protect them from the fallout of an algorithmic failure or a copyright dispute.
The absence of a "bellwether case"—a landmark judicial decision that sets a clear precedent for how AI exclusions will be interpreted by courts—remains a significant source of uncertainty. Until such a case arrives, insurers will likely continue to use the broadest language possible to protect themselves, while policyholders will be forced to look toward other avenues of risk mitigation, such as specialized AI insurance products or enhanced internal controls.
Furthermore, state regulators will play a pivotal role. Insurance commissioners are tasked with ensuring that exclusions are not "unfairly discriminatory" and that they do not leave policyholders with illusory coverage. As more filings reach state desks, we may see a patchwork of regulations where certain AI exclusions are permitted in some states but rejected or modified in others.
In conclusion, the insurance industry is currently recalibrating its relationship with artificial intelligence. Driven by a massive spike in litigation and the need for underwriting stability, the transition from broad coverage to specific exclusions is moving at a pace rarely seen in the financial services sector. For the modern enterprise, navigating this new reality will require a proactive approach to insurance archaeology, a deep understanding of policy language, and a robust commitment to AI governance to ensure that the risks of the digital age do not become uninsurable.
