Law360 Canada (September 18, 2026, 2:50 PM EDT) --
 |
| Deepshikha Dutt |
 |
| Leo Rebello |
What happens when professional judgment is no longer entirely human? Artificial intelligence (AI), once viewed as a farfetched idea, has become increasingly embedded in boardroom strategy.
As corporations expand their use of both human and artificial intelligence, the associated liability exposures are evolving as well. Claims, regulatory scrutiny and coverage disputes are emerging in circumstances that many management liability programs were not originally designed to address. While AI-related exclusions have historically appeared in general liability and technology errors and omissions (E&O) forms, the nature of the underlying risk is shifting.
AI systems now inform or influence board-level decisions, professional judgment and regulatory disclosures. Functions once driven primarily by an individual’s expertise and judgment may now incorporate input from various AI platforms. As a result, traditional directors and officers (D&O) policy language needs to account for AI’s growing role in decision-making, as well as the corporate governance structures surrounding its use.
The rise of AI
Blue Planet Studio: ISTOCKPHOTO.COM
AI has swiftly expanded from a developing technology into a conventional business tool. According to Bain & Company, 95 per cent of U.S. companies are now using AI in some capacity (“
Survey: Generative AI’s Uptake Is Unprecedented Despite Roadblocks”). In Canada, adoption is following a similar trajectory: according to KPMG Canada’s 2025 Generative AI Adoption Index, 93 per cent of Canadian business leaders report that their organizations are using generative AI in some form, up from 61 per cent the prior year (“
Canadian businesses adopting AI, but few are seeing ROI”). KPMG further confirms that 51 per cent of Canadian employees now use generative AI at work, up from 46 per cent in 2024 and just 22 per cent in 2023 (“
Beyond AI adoption: Turning Canada’s AI momentum into measurable returns”).
AI is increasingly embedded in everyday business processes, including document creation, customer service, human resources and procurement. Across sectors, organizations are using generative AI to improve efficiency, automate workflows and reduce operational costs. As such, AI adoption is no longer a differentiator; it has become an expected feature of modern business operations.
Risks in the D&O and E&O space
The dangers posed by AI to D&O and E&O insurers differ in important respects from those associated with more traditional lines of insurance. In the D&O space, exposure will often stem not from the technology itself, but from the decisions made about its use. Organizations may face scrutiny where AI systems malfunction, boards fail to exercise prudent oversight and public disclosures overstate AI capabilities. In each case, the focus of potential liability rests less on the underlying technology than on the judgment exercised in deploying, governing and supervising its use.
Similar concerns arise in the E&O context. As AI tools become increasingly integrated in professional services, errors may result not from a lack of technical expertise, but from overreliance on AI-generated outputs. A professional who accepts AI-generated research at face value, fails to identify inaccuracies, or incorporates fabricated information into a work product may face allegations that the applicable standard of care has not been met.
These exposures warrant particular attention from insurers. Governance failures may expose a company to multiple forms of liability simultaneously, while professional reliance on flawed AI-generated outputs can give rise to negligence claims. The underwriting challenge is further complicated by the potential for correlated losses across insureds that rely on the same technologies or adopt similar implementation practices. Recent incidents involving inaccurate AI-generated content and AI-enabled fraud demonstrate the broad range of risks associated with these systems and the uncertainty surrounding the manner in which such risks may materialize.
Understanding AI exclusions
AI exclusions are policy provisions that limit or eliminate coverage for losses arising from, relating to or involving the use of artificial intelligence. Depending on their wording, they may operate narrowly, excluding only losses directly caused by AI, or more liberally, excluding claims in which AI played any part in the underlying events.
In the United States, the Insurance Services Office (now known as Verisk), a leading insurance advisory organization responsible for developing standardized policy forms and underwriting tools, has released three optional endorsements effective Jan. 1, 2026, designed to remove both the duty to defend and the duty to indemnify policyholders for losses connected to AI (“
The Rise of AI Exclusions in Insurance Policies,” The Anderson Edge). Verisk anticipates filing similar exclusions for additional lines of business at a future date, signalling that the current general liability and commercial umbrella filings are only the beginning.
Canadian insurers are watching these developments closely, and similar endorsements are expected to appear in the Canadian market. Beyond standardized forms, individual carriers are drafting endorsements that go further. The
Financial Times reports that a number of major insurers have filed for regulatory approval to introduce sweeping exclusions covering any use, or even alleged use, of AI (“Insurers retreat from AI cover as risk of multibillion-dollar claims mount”). Under the most expansive carrier-specific forms, an insurer could deny a claim that it would have covered in 2025 if AI were even minimally involved in any part of the process leading to the loss (“The Rise of AI Exclusions in Insurance Policies”). Some insurers have also elected to address AI exposure through bespoke endorsements that expressly allocate AI-related risk under the policy (“The Rise of AI Exclusions in Insurance Policies”).
D&O professional liability exposure for insurers and insureds
Directors and officers face a growing range of AI-related claims. Boards that fail to implement appropriate oversight, require reporting on AI risks, or understand the implications of AI deployments may face breach-of-fiduciary-duty or derivative claims. Further, as AI becomes a material operational risk, the failure to establish AI-specific governance mechanisms becomes a potential basis for derivative claims.
“AI-washing” represents a second exposure (“
AI-Washing In Insurance: Are Companies Overpromising and Underdelivering?,”
Forbes). Companies that overstate their AI capabilities in public filings or investor presentations face securities fraud claims when the AI fails to deliver promised results. Regulators have signalled heightened scrutiny of AI-related disclosures, increasing the potential for shareholder claims.
A third category arises when AI-driven decisions go wrong (“The Rise of AI Exclusions in Insurance Policies”). Whether an AI-powered trading algorithm triggers unexpected losses, an AI credit-scoring model produces discriminatory outcomes, or an AI-driven strategic decision destroys shareholder value, the question for the insurer is whether the existing policy covers the claim, whether an AI exclusion applies, and whether the causal nexus between AI and the loss suffices to trigger the exclusion.
E&O professional liability exposure for insurers and insureds
Professionals who rely on AI face a distinct, but equally substantial, set of exposures. The most widely discussed risk is the AI hallucination, a phenomenon in which generative AI produces fabricated content, including fictitious case citations, nonexistent regulatory provisions or inaccurate financial data. When a lawyer submits a brief containing hallucinated citations, when an accountant relies on AI-generated figures that prove wrong, or when a financial adviser provides recommendations based on erroneous AI outputs, the result is a classic malpractice claim.
The duty of competence requires professionals to understand the tools they use and to verify AI-generated work product before delivering it to clients. Unreviewed AI output containing errors is, from the client’s perspective, indistinguishable from negligent work. E&O insurers must consider whether claims rooting from AI errors fall within existing coverage, whether new exclusionary language is needed, and how to distinguish between competent use of AI, which should remain covered, and reckless reliance that represents a qualitatively different risk.
Conclusion and key takeaways
The question surrounding AI exclusions is not whether insurers will address AI-related risks, but the extent to which they will do so. Where policies provide silent coverage for AI-related exposures, coverage disputes may follow. At the same time, newly developed exclusions are likely to face judicial scrutiny in an area of law where courts have traditionally interpreted coverage broadly (
Reid Crowther & Partners Ltd. v. Simcoe & Erie General Insurance Co., [1993] 1 S.C.R. 252 at para. 3;
Canadian National Railway Co. v. Royal and Sun Alliance Insurance Co. of Canada, 2008 SCC 66 at para. 32).
As AI becomes increasingly integrated into corporate decision-making and professional services, insureds and insurers should keep the following considerations in mind:
- AI exposure extends beyond technology risk. For D&O coverage, liability may arise from how boards oversee, govern and disclose the use of AI. For E&O coverage, exposure may arise when professionals do not appropriately review or verify AI-generated work.
- The wording of AI exclusions matters. Broad exclusions that capture any claim involving AI could have significant implications for coverage and may face judicial scrutiny. Clear language establishing the required connection between AI and the loss will be critical.
- Insurers should assess AI exposure at the underwriting stage. Understanding how an insured uses AI, the governance measures it has implemented and the extent of human oversight may help insurers evaluate and price the risk.
- Insureds should understand how AI may affect their coverage. Organizations and professionals should carefully review their existing and renewal policies to determine whether AI-related risks are covered, excluded or subject to specific conditions.
Deepshikha Dutt is a partner in Dentons Canada’s litigation and dispute resolution group. She also co-leads the national professional liability subgroup. For more information about Deepshikha, please visit: https://www.dentons.com/en/deepshikha-dutt.
Leo Rebello is an associate in Dentons Canada’s litigation and dispute resolution group. He is developing a broad litigation practice with an emphasis on product liability, insurance coverage and professional liability. For more information about Leo, please visit: https://www.dentons.com/en/leo-rebello.
The opinions expressed are those of the author and do not reflect the views of the author’s firm, its clients, Law360 Canada, LexisNexis Canada or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.
Interested in writing for us? To learn more about how you can add your voice to Law360 Canada, contact Analysis Editor Yvette Trancoso at Yvette.Trancoso-barrett@lexisnexis.ca or call 905-415-5811.