AI Exclusions and E&O for Engineering Reports (2026)

Since January 2026, insurers have had standard ISO endorsements that exclude generative-AI-related claims from commercial general liability policies (CG 40 47, CG 40 48 and CG 35 08), and some carriers have gone further with AI exclusions on errors and omissions policies. Whether your firm's professional liability policy covers a report drafted with AI assistance depends on your policy's wording and any endorsements on it, so the only reliable answer comes from your broker. This page explains what has changed, with sources, and the questions to ask.

Nothing here is insurance or legal advice, and nothing here says any policy covers or excludes your work. It is a summary of published reporting as of the date checked, written for engineers who sign reports.

What are the ISO generative AI exclusions?

Verisk's ISO, the insurance standard-forms organization, filed three generative AI endorsements for commercial general liability (CGL) as part of a multistate general liability filing. As described in Insurance Journal's reporting of Verisk's descriptions:

• CG 40 47 excludes bodily injury, property damage, and personal and advertising injury arising out of generative artificial intelligence.
• CG 40 48 excludes only personal and advertising injury arising out of generative artificial intelligence.
• CG 35 08 excludes bodily injury and property damage arising out of generative artificial intelligence under the products and completed operations coverage part.

Insurance Business reports that CG 40 47 (edition 01 26) became available for carriers to attach to CGL policies in January 2026, that W.R. Berkley, Chubb, Travelers, Berkshire Hathaway and AIG had each filed to adopt the ISO forms or their own equivalents by April, and that a nationwide review found 2,369 generative AI exclusion records in force across 49 states by mid-2026. These are endorsements an insurer may attach, not automatic changes to every policy.

Do they apply to my professional liability policy?

Not directly. The ISO forms above are written for the CGL policy. Engineering and architecture firms usually carry professional liability (E&O) as a separate policy, and that is the one that responds to a claim that a report was wrong. But professional liability carriers are writing their own exclusions:

• Hunton Andrews Kurth reported in May 2025 that Berkley had introduced what it called an "absolute" AI exclusion for its D&O, E&O and fiduciary liability products, barring coverage for claims based on or arising from any actual or alleged use, deployment or development of artificial intelligence.
• Insurance Business reports that AI-specific E&O exclusions are appearing more broadly across the professional liability market.
• In Insurance Journal's July 2026 reporting, a source quoted expected E&O and D&O to be where AI exclusions expand most aggressively.

How broadly an exclusion reads, and whether it reaches a report an engineer reviewed and signed, is a question of the exact wording and, if it comes to it, of a court. Do not assume either way.

What should I ask my broker?

• Does our professional liability policy, or any endorsement on it, contain an AI, generative AI or "artificial intelligence" exclusion? Please send the endorsement wording.
• Is one being added at renewal? Will the renewal application ask how we use AI?
• If we use software that drafts a report from our evidence, and our licensed engineer reviews, edits and signs it, is a claim on that report treated differently from one on a report typed in Word?
• Does it matter whether the AI tool is a general chatbot or software built for the report type with a named reviewer step?
• Do our client contracts or lender engagement terms require us to disclose AI use, and does the policy require it?
• If an exclusion applies, is an affirmative AI coverage endorsement or a standalone policy available, and at what cost?
• Does our general liability policy carry CG 40 47, CG 40 48, CG 35 08 or an equivalent?

Ask for answers in writing and keep them with the policy.

What should I give my broker?

A plain description of what the software does and does not do. The questions an underwriter asks are practical: who forms the opinions, who signs, what the tool is allowed to write, how errors are caught, and what the client is told.

For Baseline PCR, the AI Disclosure is written for exactly that reader. In short: the software drafts from the evidence your firm uploads; it does not visit the property, form engineering opinions or sign anything; every fact in the draft carries a source tag (observed, reported, documented or not provided); the cost tables are computed from your firm's unit costs, not written by the model; the reviewing engineer edits and approves, and nothing is issued without that approval; and every issued PDF states that it was drafted with AI assistance and reviewed by the named reviewer.

Does disclosing AI use help or hurt?

That is a question for your broker and counsel, not for software. Hunton's advice to policyholders is to watch for AI provisions added to existing, new or renewing policies and for AI questions on applications.

Where Baseline PCR fits

Baseline PCR is AI-assisted drafting software for ASTM E2018-24 property condition reports. The reviewing engineer is the professional of record. It does not claim, and cannot know, whether your policy covers AI-assisted drafting; confirm that with your broker before relying on it. To see what the reviewer is signing, read the 65-page sample report, and for how a general chatbot compares, see ChatGPT for property condition reports. You can run one job free before you decide.