AI D&O Exposure
The AI D&O Trap Door: Why Standard Policies Are Silent and Your Board Is Exposed
AI D&O exposure is not just a theory anymore. Rather, it is active, expensive, and accelerating. Artificial intelligence has moved from an IT function to a board-level governance issue. In just the first half of 2025, average securities class action settlements were found to be around $56 million, with about 7% new filings tied back to AI disputes.
Despite these figures, some directors assume their standard D&O program will respond to management’s mistakes, which is a risky assumption. New liability triggers are created by algorithmic bias, unpredictable outputs, and fast-changing regulations. Check your policy. If it is silent or overly broad in its exclusions, your board might be standing over a trap door.
The Litigation Shift and Growing AI D&O Exposure
AI litigation nature has changed. Earlier, claims focused on “AI-washing,” where companies overstated their capabilities. However, now plaintiffs are targeting the opposite problem: understating or poorly disclosing risk.
An example that was widely discussed involved Reddit and the impact of AI-driven search summaries rolled out by Google. It was signaled by the management that new search features would not materially impact traffic. However, as soon as the user visited and engagement dropped, a securities suit followed.
It is clear; boards must specifically assess not only how they deploy AI internally, but how competitors, vendors, and platforms use AI in ways that impact revenue. If your disclosures linger, plaintiffs’ firms will argue the gap was significant.
That is how your AI D&O exposure becomes concrete. It is not just about what your company builds but also what your ecosystem changes.
Silent Policies and the Absolute AI Exclusion
The reaction from insurance markets has been quiet, but firm. At renewal, many traditional carriers now add “Absolute AI Exclusions.” Such provisions exclude coverage for claims “based upon, arising out of, or in any way involving” artificial intelligence.
That wording is intentionally broad.
Take a practical scenario, for example. Your legal team uses an AI tool to refine an SEC filing, which leads to a securities claim later alleging misleading disclosure. The carrier may find this to be a red flag and argue that the claim “involves” AI, even if the role of the tool was minor. This could also lead to a full denial of defense.
If your D&O policy has no remarks on AI, courts and carriers may interpret that against you. Overboard exclusions, on the other hand, may create a coverage trap door that only opens when litigation hits.
If left unchecked, AI D&O exposure is likely to reveal itself at the worst possible moment: after counsel is retained and invoices begin to accumulate.
Carriers Building Around AI Instead of Avoiding It
Not all markets are retreating; some carriers are building promising solutions designed to address AI D&O exposure directly.
Here are three that stand out:
- Relm Insurance
- Coalition
- Corix
Practical Steps to Reduce AI D&O Exposure in 2026
The best call for boards is to act now, not at renewal.
- Silent AI audit
Ask directly whether the D&O form mentions AI explicitly. In today’s market, no reference often signals a coverage gap.
- Review Pay-on-Behalf Provisions
Confirm whether defense costs are paid directly by the carrier. Reimbursement structures can strain liquidity, especially for startups.
- Identify Absolute Exclusion Language
Quickly scan for wording that excludes anything tied to machine learning, automated systems, or algorithmic processes. Push for narrower, manuscriptal language where possible.
If your company heavily relies on AI, your exposure is not hypothetical, and the risk is high – as most standard programs have not been built with this environment in mind. The risk is not specific to just D&O; it further splits into cyber, E&O, and broader commercial lines as exclusions expand, and plaintiffs test new theories.
Closing the Trap Door
The insurance market is splitting into two camps – where one avoids AI through silence and sweeping exclusions; the other builds structured solutions that acknowledge how businesses operate.
When the next settlement wave arrives, boards that treat AI D&O exposure as a governance priority will fare better.
The solution to wanting a direct review of your current D&O structure is as simple as asking. A focused policy analysis can identify hidden exclusions, weak defense triggers, and disclosure vulnerabilities before they become public problems.
