Coverage layer
AI DIC Excess from Mayflower Specialty is a difference-in-conditions layer that sits over a company's existing D&O, EPL and E&O policies and adds affirmative AI coverage where those policies are silent on AI or exclude it. It lets a company address the AI gap without replacing the program and insurers it already has.
At a glance
Definition
Definition
A difference-in-conditions (DIC) policy is written to respond where an underlying policy's terms are narrower than its own, for example because the underlying policy excludes a type of claim, so it fills gaps in cover rather than only adding limit.
DIC cover is familiar from property insurance and from Side A D&O programs, where a DIC layer responds when an underlying policy will not. AI DIC Excess applies the same idea to AI: it sits over the D&O, EPL and E&O program a company already carries and is written to respond where those policies are silent on AI or exclude it.
It suits a company whose program is otherwise sound apart from its AI wording, because it adds affirmative AI coverage without moving the program to new insurers. How the layer attaches, and over which underlying policies, is set out in its terms and confirmed in underwriting.
Real case
ReportedMay 2025
Hunton Andrews Kurth reported an exclusion introduced by W. R. Berkley that removes cover for claims “based upon, arising out of, or attributable to” any actual or alleged use, deployment or development of AI by any person or entity. An exclusion of that breadth is the situation a DIC layer is designed for.
How it works
AI DIC Excess sits over the whole program rather than inside any one tower. What it does for a given AI claim depends on how the underlying policy treats that claim, as the three cases below show.
Choose what the underlying policy does
If the D&O, EPL or E&O policy covers the AI claim, it responds first, as it would today. The AI DIC Excess layer stays in place above the program.
If an AI exclusion removes cover for the claim, AI DIC Excess is written to respond in place of the excluded cover, subject to its terms.
If the policy neither covers nor excludes AI, the answer is left to a coverage dispute. AI DIC Excess is written to respond affirmatively where the underlying policy does not, subject to its terms.
The diagram is a simplified illustration. How the layer attaches, and over which underlying policies, is set out in its terms and confirmed in underwriting.
Written for
AI DIC Excess is written for AI claims that fall into gaps in an existing D&O, EPL and E&O program. Those gaps usually open in one of four ways.
An insurer adds an AI exclusion to the D&O, EPL or E&O policy at renewal, and claims the program used to answer now fall outside it. The rest of the program is unchanged, which is why a layer that fills the AI gap can be simpler than moving every line.
The underlying policy neither covers nor excludes AI, so whether it responds to an AI claim is left to argument after the loss. A DIC layer is written to respond affirmatively in that gap, subject to its terms.
An AI claim can carry features of more than one line, such as an HR technology failure that is partly an employment claim and partly a professional services claim. Each policy may then point to the other, and a DIC layer over both lines is written to respond in that gap, subject to its terms.
A program built from several insurers can carry a different AI definition or exclusion in each policy. A single DIC layer over the program is designed to address AI consistently across them, subject to its terms.
Whether a particular claim is covered depends on the policy wording and the facts. AI DIC Excess is written for these gaps, subject to the policy terms.
The market
Insurers are moving from silence to explicit AI wording, as the sourced reports below show. They describe the market rather than any one company's policy, which is why each renewal in a program should be read for new AI definitions and exclusions.
November 24th 2025
ReportedThe Financial Times reported that AIG, Great American and W. R. Berkley had sought US regulatory approval for AI exclusions in their policies.
Source Insurance Business, reporting the Financial Times, November 24th 2025 (opens in a new tab)
January 2026
In effectVerisk's ISO generative AI exclusion endorsements for commercial general liability, CG 40 47, CG 40 48 and CG 35 08, took effect in January 2026, giving insurers standard wording to remove generative AI exposures from CGL policies.
Source Big “I” Virtual University summary of the Verisk filing (opens in a new tab)
September 16th 2026
ReportedInsurance Business reported that W. R. Berkley, Chubb, Travelers, Berkshire Hathaway and AIG had filed to adopt the ISO generative AI forms or their own equivalents by April 2026.
Source Insurance Business, September 16th 2026 (opens in a new tab)
Choosing a structure
Both structures add affirmative AI coverage. The right one depends mainly on how many lines have AI gaps, how settled the existing program is and when it renews.
| Consideration | AI DIC Excess | A primary module (AI-D&O, AI-EPL or AI-E&O) |
|---|---|---|
| Existing program | Keeps the current D&O, EPL and E&O policies and insurers in place | Adds a policy written for AI claims in one line, alongside the existing program |
| Where the gap is | Suits gaps in several lines, or a program where AI wording differs from one insurer to the next | Suits a gap concentrated in one line, such as AI hiring tools under EPL |
| What changed | Suits a renewal that has added an AI exclusion to a program that is otherwise sound | Suits a company that wants named AI cover in a line whatever its current wording says |
| Timing | Worth raising as soon as a renewal shows new AI wording, so the layer can be in place when it takes effect | Worth raising when that line is being placed or renewed |
Cost depends on underwriting for either structure, so the two can only be compared once quotes are in hand. Both are written on a claims made and reported basis and underwritten on the applicant's AI governance.
Scenarios
These scenarios show the gaps a DIC layer is written for: an exclusion at renewal, a claim between lines and a declined claim.
These scenarios are hypothetical, and whether a policy responds depends on its wording and the facts.
An absolute AI exclusion is added to a company's D&O policy at renewal. Six months later, shareholders sue the directors over statements about the company's AI, and the D&O insurer points to the exclusion.
Hypothetical · Written for: AI DIC Excess
An HR software company's AI scheduling tool is blamed for disability discrimination against a customer's staff. The company's E&O insurer says the claim is an employment matter, and its EPL insurer says the people affected were never its employees.
Hypothetical · Written for: AI DIC Excess
A clinic group's EPL policy excludes claims arising from automated decision-making. Applicants allege that its AI interview tool screened out candidates with speech impairments, and the EPL insurer declines the claim under that exclusion.
Hypothetical · Written for: AI DIC Excess
Who it is for
AI DIC Excess is designed for companies that already carry D&O, EPL and E&O cover and want its AI gaps closed without rebuilding the program.
The supplemental application also places these uses of AI outside Mayflower's underwriting appetite:
Underwriting
AI DIC Excess is underwritten over the program it sits on, so underwriters need to see that program as well as the company's AI governance. A copy of each underlying policy's AI definitions and exclusions helps the review. The questions below are plain-language summaries of the supplemental application, which has the exact wording.
Completing the application does not bind coverage. How underwriters assess AI risk
Section IX · Prior and current insurance
For each current or expiring D&O, EPL and E&O policy, who is the insurer, and what are the policy period, limit, retention and premium?
Why it is asked: The DIC layer is written over these policies, so underwriting needs the whole program.
Section IX · Prior and current insurance
What retroactive date is proposed, and when does the current insurer's policy expire?
Why it is asked: Claims made and reported cover depends on dates lining up across the program.
Section VIII · Claims and loss history
Has any insurer declined, canceled or non-renewed coverage for the company?
Why it is asked: Explains how the current program came to look the way it does.
Section VIII · Claims and loss history
Is the company aware of any facts or circumstances that could reasonably give rise to a claim?
Why it is asked: Known circumstances are identified before cover starts rather than disputed afterward.
Section III · AI governance
Has the company adopted a formal AI governance framework, and how fully is it implemented across production AI systems?
Why it is asked: Governance is the basis of underwriting for every Mayflower module, including the DIC layer.
Section II · AI systems overview
How many distinct AI systems are in production, and what is the highest level of integration and decision materiality among them?
Why it is asked: Sizes the AI exposure that the underlying program may no longer cover.
FAQ
Further reading
AI DIC Excess
Apply online, or talk to the team about how AI DIC Excess would sit over your current D&O, EPL and E&O policies.
Placed through brokers on A- (Excellent) AM Best rated paper.
