Once the domain of science fiction, artificial intelligence (AI) has become an inescapable reality that has rapidly
reshaped the corporate landscape. The risks for companies and their directors and officers (D&Os) have evolved just
as quickly. Yet, despite growing attention on AI-related litigation and a narrative from some commentators that
insurers are responding by broadly excluding AI-related risks from D&O coverage, the reality is far less dramatic.
Robust D&O protection for AI-related exposures remains widely available, even as plaintiffs test novel theories of
liability.
AI’s Growing Role in Corporate Valuations
A staggering 337 of the 498 S&P 500 companies that held earnings calls between March 15 through June 11 of this
year—68% of them—discussed AI on the calls: more than double the 5-year average (164), and more than triple the
10-year average (103).1
Equally notable is that, from an average stock-price perspective, companies discussing AI on recent earnings calls
have materially outperformed those that did not—since March 31, 2026 (average increases of 12.7% vs. 2.6%) and since
December 31, 2025 (average increases of 13.1% vs. 7.7%).2
Consistent with the adage that “the bigger they are, the harder they fall,” however, higher valuations can mean
larger alleged damages. This has been especially true for AI-related securities class actions, which accounted for
only 13% of the securities class actions filed in the first half of 2026, but an outsized 73% of the alleged
investor losses in all such actions.3
The Scale and Complexity of AI Financing Are Compounding Risks
Current AI securities litigation may reflect little more than market volatility generated by AI’s early hype. A more
significant wave of claims could emerge later if massive AI investments fail to produce long-term revenues.
The AI arms race is being funded at a scale rarely seen outside of wartime mobilization or national infrastructure
projects. Leading hyperscalers are estimated to have spent $300–$400 billion on AI infrastructure in 2025 alone.
Adjusted for inflation, that is more than the U.S. government spent on the entire 13-year Apollo moon landing
program. And AI infrastructure buildouts are projected to require in excess of $5 trillion by the end of the decade.4
Although estimates vary, it is widely understood that AI products and services do not currently provide and are not
projected in the short term to provide revenues even remotely commensurate with these massive AI-related capital
expenditures.
Exacerbating the risk is that companies are increasingly relying on complex alternative financing arrangements to
meet these funding requirements. These arrangements might expose companies and their D&Os to investor lawsuits
alleging inadequate disclosures concerning such alternative financing structures, consistent with multiple U.S.
Senators’ warnings that the “convoluted and opaque” financing structures may enable companies “to obscure the true
nature of their balance sheets” and “appear healthier and less leveraged than” they actually are.5
How AI-Related Litigation is Evolving
Whether AI buildout investments ultimately produce sustainable returns remains to be seen. What is already apparent,
however, is that the AI-related D&O litigation landscape is evolving.
Just two years ago, the initial wave of AI-related D&O claims generally involved “AI washing,” that is, companies
making allegedly false and misleading representations concerning their AI capabilities. Although AI washing claims
are still being brought, the plaintiffs’ bar is increasingly testing more nuanced theories of D&O liability that can
be distilled into two groups: (1) “failure to monitor” shareholder derivative claims; and (2) shareholder securities
class actions alleging that defendants concealed AI-related risks.
- Shareholder derivative claims alleging “failure to monitor”: Aon predicted two years ago that,
when companies are accused of intellectual property infringement in connection with their AI practices,
plaintiffs will bring follow-on shareholder derivative suits repackaging those allegations into claims against
the companies’ D&Os for failing to monitor the infringement in breach of their fiduciary duties of loyalty.
There are now at least four such derivative lawsuits pending against technology companies.6
- Securities class action claims alleging inadequate risk disclosures: In contrast to AI washing
cases alleging the defendants overstated their AI capabilities, plaintiffs have also pursued a distinct type of
AI-related disclosure claim rooted in defendants allegedly understating their AI risks. Plaintiffs have brought
the latter types of claims against numerous companies and their respective D&Os.
For example, a large technology company and certain of its D&Os are facing securities class action claims alleging that the defendants promoted the company's AI-related growth prospects while misleadingly concealing the scale of capital expenditures required for its AI infrastructure buildout. The claims further allege that the defendants failed to disclose the resulting impact on the company's balance sheet and broader risk profile.
It is reasonable to assume that, as the proliferation of AI continues, the D&O claims environment will continue to
evolve. As just one example, Aon anticipates shareholder litigation challenging the allegedly “opaque” financing
arrangements companies are using to fund the AI infrastructure buildout, described above.
D&O Insurance Continues to Respond
Generally, D&O policies continue to provide broad coverage for AI-related exposures. More precisely, D&O
policies’ insuring clauses and key definitions generally are agnostic about whether a given Claim or Loss arises
from AI, and Aon has not seen insurers attempt to impose AI-specific exclusions on D&O policies. While AI is
unprecedented in many ways, and plaintiffs’ lawyers are branching out from mere “AI washing” theories in D&O
lawsuits, neither the novelty of the technology nor the variation of allegations from one AI case to the next
necessarily alters the fundamental nature of the claims themselves.
Subject, of course, to the terms, conditions, exclusions, and applicable law governing any particular policy, at
least with respect to public companies, AI-related D&O litigation has thus far largely fallen, and is likely to
continue falling, within traditional categories of shareholder litigation core to D&O coverage, namely,
disclosure-based securities class actions and shareholder derivative suits.
Some commentators have suggested that the filing of AI-related exclusions by certain insurers may signal a broader
shift in the market. Although Aon agrees that insureds and brokers must be alert for any such exclusions, and it is
conceivable that insurers have sought to or might soon add such exclusions onto D&O policies (e.g., for niche
risks), Aon has not observed the adoption of such exclusions as a widespread feature of D&O placements.
Likewise, while some have questioned whether D&O insurers actually contemplate assuming AI-related exposures,
the fact of the matter is that the AI D&O claims to date generally arise from the very types of shareholder
allegations that D&O insurance has long been designed to address.
Indeed, insurance underwriters’ growing scrutiny of AI governance, controls, oversight, and disclosure practices
during underwriting meetings suggests not that insurers regard these risks as foreign to D&O insurance, but
rather that they view them as emerging manifestations of the same management liability exposures the product has
always been intended to cover.
Why AI Governance Matters More Than Ever
Strong AI governance is an essential complement to D&O coverage. Beyond securing appropriate D&O coverage,
companies should be prepared to demonstrate that AI risks are governed with the same discipline applied to
cybersecurity, privacy, and financial controls.
Best practice governance should include documented board oversight, clear executive accountability—whether through a
Head of AI or a senior C-suite leader with stated oversight duties—formal AI policies and standards rolled out to
employees, risk assessments and approval workflows, and ongoing monitoring through testing, audits, KPIs, incident
management, training, and documented remediation.
This governance record is becoming increasingly important as proxy advisors, including ISS and Glass Lewis, focus
more closely on how public companies disclose and oversee AI-related risks for investors. The same governance record
likewise can be critical in the defense of investor claims, such as AI-related shareholder derivative suits alleging
directors’ and officers’ failures to monitor AI-related risks, as alluded to above. This is especially true when
AI-related risks are “mission critical” to companies, given that D&O oversight of mission critical risks has
drawn particularly heightened scrutiny from courts in recent years.