Global Directors’ and Officers’ Survey Report 2026 – AI

From buzzword to boardroom peril: AI climbs the D&O risk rankings

The results of Global Directors’ and Officers’ (D&O) Liability Insurance Survey 2026 paint a striking picture: artificial intelligence (AI) and machine learning risks are climbing the boardroom agenda faster than almost any other category of risk. In a year that has seen an unprecedented acceleration of AI deployment across industries and a corresponding increase in regulatory, ethical and litigatious scrutiny, it is perhaps unsurprising that 56% of global respondents to the survey now rate AI and machine learning as "very important" or "extremely important" issues for their directors and officers.

What is remarkable is the trajectory: that figure stood at just 48% in 2024 and 51% in 2025. We are seeing a consistent year-on-year climb. Indeed, AI was one of only two risk categories in the entire 2026 survey to register an increase against the prior year (up five percentage points), at a time when the majority of other risk categories declined.

For those of us who advise on insurance and risk management, these numbers demand attention. The question is no longer whether AI will generate D&O exposures, but how quickly those exposures will crystallise and whether current insurance programmes are ready.

The global picture: AI moves up the rankings

Global D&O Liability Insurance Survey 2026 confirms that AI and machine learning has firmly established itself among the top-tier risks facing directors and officers worldwide. When risks are categorised thematically, AI sits alongside data loss and cyber attack within the "cyber-related risks" grouping. However, AI risk is a distinct category in its own right, encompassing governance failures, strategic missteps, under-performance of models, regulatory non-compliance and reputational harm that has the potential to raise different issues, cause different losses and therefore requires different expertise and risk assessment.

When respondents who consider AI risk to be important were asked what specifically concerns them most, the answers revealed a nuanced and layered risk landscape:

AI related concerns

Source: Global D&O Liability Insurance Survey 2026

These results pose a challenging landscape for boards. On the one hand, the top concern, AI-generated errors and misinformation, reflects anxiety about the reliability of AI outputs and the reputational and legal consequences when AI gets things wrong. On the other hand, the third-ranked concern, strategic failure to adopt AI, reveals the opposite fear: that companies which fail to integrate AI swiftly enough will fall behind competitors, exposing directors to allegations of strategic mismanagement. A further result worthy of comment is, while 50% of respondents identified errors and misinformation as a concern, only 32% considered weak governance and controlled AI use as concern. Perhaps this suggests that the focus on AI governance is lagging behind AI deployment. Recent trends in regulatory pressure and enforcement, as well as the evolution of US securities class actions, suggest that there are already real consequences materialitiesfor directors who fail to engage meaningfully with AI oversight.

The North America outlier: Why AI risk resonates so deeply across the Atlantic

Perhaps the most prominent finding in the 2026 survey is the extent to which North American respondents are concerned about AI and machine learning risk as compared to their global counterparts. At 71%, the proportion of North American respondents rating AI as "very important" or "extremely important" is notably higher than the global average of 56%. Interesting only the Middle East and Africa also had a proportion higher than 60% who considered AI as “very important” or “extremely important”. North America is the only region where AI and machine learning broke into the top seven risks for directors and officers, ranking fifth at 71%.

AI/machine learning rating by region

Source: Global D&O Liability Insurance Survey 2026

While Africa (74%) and the Middle East (74%) recorded higher absolute percentages for AI risk than North America, those regions' overall risk scores were elevated across almost all categories, meaning that AI did not stand out sufficiently to rank within their top seven concerns. In North America, by contrast, AI was singled out as a priority concern relative to other risks. This would appear to speak to the intensity and pervasiveness of AI discourse in the United States.

Several factors are likely at play. The United States is a global centre of generative AI development. Companies such as OpenAI, Google DeepMind, Anthropic and Meta have driven a pace of innovation that has placed AI squarely in the public consciousness in the US, from congressional hearings on AI safety to high-profile litigation over copyright, deepfakes and algorithmic bias. 

The regulatory landscape adds further complexity. While the European Union has pursued a prescriptive approach, the United States has adopted a more fragmented, sector-specific posture. Executive orders, Federal Trade Commission enforcement actions, and state-level AI legislation create a patchwork of obligations that North American directors must navigate. This regulatory fragmentation creates heightened compliance risk for multinational organizations, particularly those deploying AI across multiple jurisdictions with differing requirements. The risk is not merely of regulatory sanction, but of the various reputational and litigious consequences that follow.

Size matters: Larger companies feel the pressure

The survey also reveals a clear correlation between company size and AI risk perception. Among companies with revenues of USD 5 billion or more, 66% of respondents rated AI as very or extremely important. By comparison, just 45% of respondents from companies with revenues below USD 50 million gave the same ranking. This gap likely reflects the greater deployment of AI tools at scale within larger organizations, the more complex governance structures required to oversee that deployment, and the heightened litigation exposure that comes with a larger public profile and deeper pockets.

AI/machine learning rating by company size

Source: Global D&O Liability Insurance Survey 2026

Across industries, finance and insurance lead the way in identifying AI as a concern at 63%, followed by healthcare at 59%, with the industrial sector at the lower end at 52%. The services sector is the only industry where AI broke into the top seven risks, ranking seventh at 56%. This is consistent with the services sector's heavy reliance on AI-driven client-facing tools, where errors in AI outputs can have immediate and visible consequences.

Are D&O policies ready?

AI-related claims may test the boundaries of traditional D&O policy wordings in several respects. For example, will "wrongful acts" definitions capture negligent oversight of an AI system? Will AI-related regulatory investigations fall within insured "proceedings"? And will insurers seek to characterise AI losses as operational or technology failures excluded under professional liability or technology E&O policies?

Global D&O Liability Insurance Survey 2026 found that only 51% of boards feel they have sufficient skills and knowledge to provide effective oversight of AI, making it the second-lowest scoring topic of the fourteen areas assessed, ahead only of climate risk and transition (48%). This knowledge gap, if left unaddressed, could itself become a source of D&O liability. Directors who cannot demonstrate a good understanding of the AI tools their organizations deploy may struggle to satisfy their duty of care and insurers may scrutinise the adequacy of board-level AI governance when assessing claims.

Looking ahead: Questions for boards and their advisers

Global D&O Liability Insurance Survey 2026 results prompt a series of challenging questions that boards, risk managers and their legal and insurance advisers should be asking now.

First, is your organization's AI governance framework sufficiently robust to withstand regulatory scrutiny and shareholder challenge and is that governance documented in a manner that demonstrates board-level engagement? Recent reports concerning the capabilities of next-generation generative AI models and the potential risks those models pose to network security illustrate the challenges boards are likely to be facing in the next 12 months. This includes risks relating to intellectual property ambiguity (where there has been minimal human input), data privacy, new variations of cyber attack taking advantage of the use of AI as a tool and vendor accountability gaps.

Second, has your D&O insurance programme (as well as your wider insurance programme) been stress-tested against AI-specific loss scenarios, including regulatory investigations, algorithmic discrimination claims, IP infringement and securities litigation arising from AI-related disclosures?

Third, in an environment where 38% of respondents cite "strategic failure to adopt AI" as a top concern, how is the board documenting decision-making on AI adoption to protect against hindsight-driven claims that they moved either too fast or too slow?

Fourth, as AI regulation continues to evolve across jurisdictions, are multinational organizations building sufficiently agile compliance programmes or are they exposed to gaps between the EU's prescriptive approach and the US's fragmented framework?

AI risk is no longer a future concern. It is a present reality that is reshaping the directors' and officers' risk landscape. For boards, the imperative is to govern AI development, adoption and deployment proactively. For insurers, the challenge is to assess where and how AI risk sits within ‘traditional’ lines of insurance and, where gaps in cover emerge, to develop products that respond meaningfully to this evolving risk. And for those of us in the business of insurance recovery, the task is to ensure that when the inevitable AI-related claims arise policyholders are positioned to secure the protection they have purchased.