AI Documentation for Carrier Renewal

Carrier Activity Snapshot

For the per-carrier breakdown of who has published AI guidance, who has filed exclusions, and who is silent, see the dedicated carrier tracker. This page covers what those carrier moves mean for renewal documentation: coverage mechanics, what carriers ask on applications, mid-policy incident handling, and the nine artifacts a firm should bring into its renewal cycle. The carrier tracker is the structured, sortable view; the rest of this page is the editorial framing.

Public versus gated model policies. Two LPL mutuals have produced model firm AI policy templates, and their distribution postures are opposite. OBLIC's Model AI Use Policy is available on request from Loss Prevention; it is not posted publicly. The Oregon PLF posts both its Generative AI Policy Development Guidelines and a companion Sample Generative AI Policy for Law Firms as free public downloads. Oregon lawyers have explicit permission to use and modify the materials. That contrast is structural: as the only mandatory bar-administered LPL program, the PLF treats the model policy as a public good, while carrier-administered programs competing for retention treat it as a member benefit. Firms outside Oregon can read the PLF materials for structural reference, but PLF's permission grant covers Oregon practitioners only. PLF text itself is not redistributable in a non-Oregon packet. This project follows the public posture: a Policy Template and a separate pre-adoption decision framework are freely available, drawing on PLF's two-document approach without reproducing PLF language.

What Carriers Are Asking

The clearest documented example is on AmTrust's Lawyers Professional Liability Application (form LPLPRO-APP-01 0523, May 2023), Section III, question 13:

"Allow the use of Artificial Intelligence software to draft documents. If checked, please attach description."

That is the only verifiable verbatim AI question on a US LPL application form located in our research as of 2026-04-29. Most carrier applications are gated behind producer portals or online apply flows. CNA's current New Business Application (form CNA105636XX 01-2024), the Berkley Insurance LPL form (10-14 revision), and the Lawyers Mutual Insurance Company of Kentucky application all lack AI questions in their public versions.

Beyond the documented AmTrust question, carrier-published guidance and on-record broker reporting indicate that underwriters are now asking firms to document specific AI controls. Stan Sterna, SVP and risk control lead at Aon, framed the underwriter posture in Digital Insurance / Accounting Today (April 21, 2026):

"They're asking firms, 'Do you use AI? Do you police it? Do you have protocols in place?'"

Sterna also notes that cyber underwriters are years ahead of professional-liability underwriters on AI, so PL renewal questions are still catching up to where cyber renewals already are. Jencap senior broker Sean Burke separately reports the same trajectory in IA Magazine (March 2026). Both accounts indicate underwriters are now asking firms to document:

• Which AI tools are in use at the firm
• Whether a written AI policy exists
• How client confidentiality is protected in AI tools
• What training attorneys and staff have received
• Whether AI-generated citations are verified before filing
• Whether AI errors are logged and reported

This list reflects the direction underwriters describe, not (yet) verbatim renewal-application questions outside the AmTrust example.

The Nine Documentation Artifacts

A firm that maintains these nine artifacts has the documentation needed for carrier conversations in 2026. The first eight are records the firm produces under ABA Opinion 512 and the malpractice-renewal cycle. The ninth is a meta-checklist for reviewing the firm's own LPL policy for AI exclusion endorsements before binding the renewal.

Coverage Mechanics: Exclusions and Disclosure

Four coverage mechanics are worth understanding before renewal.

As of May 2026, no major US LPL carrier has publicly attached an AI-specific exclusion to its named lawyers' professional liability form. Bar-affiliated mutuals (ALPS, OBLIC, TLIE, ISBA Mutual, Lawyers Mutual NC, FLMIC, MLM, LMICK, WILMIC, ALAS, LMIC California) publish risk-management guidance, not form exclusions. Multiple independent sources have documented unnamed LPL/E&O programs filing AI-related endorsements quietly through 2024 and 2025. Three sources are on record. First, an October 2024 Vanguard Specialty Insurance memo by Stephen Van Wert describes a manuscript exclusion on an unnamed LPL program (memo not publicly hosted; cited via specialty broker reporting). Second, an L Squared Insurance Agency blog post describes a "non-standard lawyers professional liability program" with this exclusion: any actual or alleged use of generative artificial intelligence by the law firm. Third, WTW's Insurance Marketplace Realities 2026 (October 2025) describes a structural break between January 2025 and January 2026. Renewals bifurcated: some carriers offered affirmative-but-conditioned AI coverage, others swept exclusions, most showed silent AI coverage. Carrier identities, form numbers, and states were not disclosed in any of these sources. Treat unnamed-program reports as a market signal, not a single carrier's policy decision.

AI exclusions in adjacent commercial lines. W.R. Berkley filed Form PC 51380 00 (06-24), "Artificial Intelligence Exclusion (Absolute)," amending its Directors & Officers (PC 51000 DO), Employment Practices (PC 51000 EPL), and Fiduciary (PC 51000 FLI) coverage parts. It covers any "actual or alleged use, deployment, or development of Artificial Intelligence," broadly defined. AI is defined as any machine-based system that "infers, from the input it receives, how to generate outputs," per the form language. On its face, the exclusion does not attach to LPL forms. Hamilton Select Insurance has reportedly applied a "Generative Artificial Intelligence Exclusion" to professional-liability policies. That exclusion is defined to cover "ChatGPT, Bard, Midjourney, or Dall-E"-type systems (verbatim quoted by Hunton/Law360, July 2025 and Zelle, October 2025; primary form not located). Verisk/ISO is rolling out three new general-liability exclusion endorsements (CG 40 47, CG 40 48, CG 35 08) effective January 2026. These are general liability forms, not LPL.

The LPL parent-group market is split, not uniformly retreating. W.R. Berkley sits on the exclusion side; AXIS Capital, parent of the AXIS PRO underwriter for the Aon Attorneys Advantage program, has co-invested in Armilla, an MGA offering affirmative AI coverage with Lloyd's/Chaucer capacity. Underwriter posture varies by parent group, not just by line.

Indirect coverage paths under existing LPL exclusions. Even without an express AI exclusion, ALPS' August 2025 article walks through how four existing LPL exclusions and definitions may already deny coverage on AI facts. First, the policy's "professional services" definition: AI tasks may not qualify if the lawyer "blindly relied" on output without validation. Second, the intentional-acts exclusion: a knowing decision not to verify a citation may be deliberate, not negligent. Third, the breach-of-confidentiality exclusion: typically carved back only for negligent breach, so pasting client data into a public AI tool may fall outside the carve-back. Fourth, the unauthorized practice of law exclusion, where AI tools interact with the public without attorney oversight.

Material misrepresentation on the application. If a renewal application asks about AI use, the answers become material representations under standard professional-liability policy terms. Material misrepresentation can void coverage retroactively. That principle applies to AI questions as to any other. The ABA Journal has separately flagged that AI-driven work may not fall within the policy definition of "professional services," creating a coverage-gap risk distinct from misrepresentation.

Firms should review renewal questionnaires and policy terms for AI-specific exclusions, sublimits, or definitional gaps before binding. Bar-opinion implications by state are covered in the State Tracker. The underlying governance document is the policy template. A renewal-ready audit pass uses the compliance checklists.

What Public Filings Say

Carriers underwriting LPL through publicly-traded parents have started conditioning investors to expect AI-driven claims and AI-driven rate or form regulation. The FY2025 10-Ks filed in February 2026 split into three patterns.

Internal-use framing only. Travelers, Chubb, and Markel describe AI as an internal productivity tool, modeling input, or competitive concern, but do not position AI as something they are underwriting on policies they sell.

Hybrid framing: regulatory exposure plus novel claims. CNA Financial, the largest US LPL writer (10-K filed February 10, 2026), warns investors of novel AI-driven exposures. Its filing states: "Policyholder use of AI could introduce novel exposures that may result in new or increased claims." CNA adds that widespread AI adoption could disrupt entire industries. Item 1A warns of GAI "hallucination," bias amplification in underwriting and claims, and AI-enabled misconduct by employees and third-party administrators. Item 1 Current Regulation lists "artificial intelligence" alongside cybersecurity and climate change as evolving regulatory areas.

Hanover Insurance Group's FY2025 10-K (filed February 20, 2026) uses similar cautionary language. Item 1A warns that AI and other new technologies "may also create unforeseen exposures or coverage issues under the policies we write or introduce new forms of claims fraud or cybercrime." See Item 1A.

Regulated-exposure framing with named regimes. W.R. Berkley's FY2025 10-K is the only filing among the public-five to name specific AI regulatory regimes. Its Item 1 Regulation discussion cites three regulatory frameworks: the NAIC Model Bulletin on AI, Colorado ECDIS Regulation 10-1-1, and NYDFS Circular Letter 7. That is consistent with W.R. Berkley being the only public US carrier to have published an Absolute AI Exclusion form (PC 51380), and aligns with the June 2024 AM Best ICR outlook revision to Positive for W.R. Berkley.

The pattern matters for LPL renewal. CNA (the largest US LPL carrier) and Hanover (a top-15 commercial writer with a substantial Specialty professional liability book) are both telling investors to expect AI-driven claims on the policies they sell. Neither has filed a named AI exclusion on their LPL forms. Documented governance closes the gap that renewal-side underwriters are being told to start pricing.

The Affirmative-Cover Side of the Same Shift

The form-side AI exclusions and the application-side AI questions are only one half of the 2025-2026 market shift. The other half is a parallel build-out of affirmative AI liability cover, designed to fill the gap that exclusions create. Several Lloyd's syndicates and US MGAs are now writing dedicated AI policies that sit alongside, not instead of, the firm's LPL.

Lloyd's syndicate Chaucer (Syndicate 1084) and MGA Armilla launched Affirmative AI Liability Insurance on April 30, 2025, with USD 25M limits. The product expanded on February 10, 2026 to "Vanguard AI," a coordinated structure bundling cyber, technology E&O, and AI liability with predefined allocation rules. Chaucer's Piers Tuggey explicitly framed it as a "conscious decision to partner rather than stretch cyber policies beyond their original intent." Armilla's marketing names lawyer-hallucination cases as in-scope.

Counterpart, a US MGA, expanded its Affirmative AI Coverage to Miscellaneous Professional Liability on November 26, 2025, adding a Tech E&O insuring agreement at the same time.

The London Market Association's International Professional Indemnity Committee published its first AI-and-E&O report on September 18, 2025. The committee recommended an exclusionary approach for carriers not intending to cover GenAI. It named lawyer hallucinations and SRA-approved Garfield.Law as risk vectors. Munich Re aiSure, Mosaic, Hiscox, AXA XL, Beazley, and Counterpart all offer affirmative AI cover in some form.

Operational implication for the firm. Documented AI governance is now two-sided evidence. It is the documentation the existing LPL underwriter wants, to avoid dropping the firm from a preferred-risk tier or attaching a manuscript AI exclusion at renewal. It is also what a dedicated affirmative AI carrier requires for the risk assessment that gates the dedicated AI policy. One audit report, one vendor file, and one incident log answers questions on both applications. A firm that documents once is positioned for either market direction.

If an AI Incident Happens Mid-Policy

Most lawyers' professional liability policies require notice of two things: a claim (defined as a demand for money or services); and circumstances the insured reasonably believes could lead to a claim. A potentially fabricated citation in a filed brief, an AI-related confidentiality breach, or a court order to show cause based on AI use may meet either threshold. Early notice is generally protective. Late notice can void coverage under most policy forms.

Material misrepresentation risk runs the other direction. If the renewal application asked about AI use, the answers become representations under standard policy terms. The documented AmTrust example (Section III, question 13) is the clearest case. If the firm's answer no longer matches actual practice, that prior representation may be material. Late disclosure surfacing during a claim investigation is the worst position for a firm.

What to prepare for the broker conversation:

• Which filing or matter the AI incident occurred in.
• Copies of the firm's AI policy as of the application date and as of the incident.
• An Incident Log entry documenting the firm's response.
• Verification Log entries (see Verification Log) demonstrating the citation-check process.
• Any court order, motion to show cause, or correspondence with opposing counsel.

Policy language controls. Read the firm's specific policy form, the application that bound it, and any endorsements; consult the broker on notice timing and the firm's ethics counsel on Rule 3.3 and Rule 8.3 questions. The Incident Response runbook walks the operational steps in order.

Sources

• ALPS, "Insurance Coverage Issues for Lawyers in the Era of Generative AI" (August 21, 2025): alpsinsurance.com
• OBLIC, "Guidance on Attorneys' Use of Generative AI": oblic.com
• TLIE, "Generative AI: Recommendations for Responsible and Ethical Usage": tlie.org
• ISBA Mutual, "Legal Ethics of AI: Adapting to Challenges with New Technology": isbamutual.com
• Lawyers Mutual NC, "Embracing AI in Your Law Practice Safely and Ethically": lawyersmutualnc.com
• LMICK, "LMICK Minute Issue #58": lmick.com
• ALAS, "You, Me, and ChatGPT": alas.com
• Wisconsin Law Journal, "Querious, Wisconsin Lawyers Mutual launch strategic relationship" (February 25, 2026): wislawjournal.com
• Markel, Sal Pollaro, "AI and professional liability: Achieving balance between rising adoption and generating risk" (January 13, 2025): markel.com
• CNA via PA Bar Insurance Program, "Practical Advice on Use of AI Ethically" (March 2025): pabarinsurance.com
• The Hartford, "Use of Artificial Intelligence" (Panel Counsel Case Management Guidelines): thehartford.com
• IA Magazine, "How Generative AI Is Reshaping Professional Liability Risk for Law Firms" (March 9, 2026): iamagazine.com
• Zelle LLP, "AI Update: The Growing Trend of AI-Related Insurance Policy Exclusions" (October 31, 2025): zellelaw.com
• ABA Journal, "Does your professional liability insurance cover AI mistakes? Don't be so sure" (2025): americanbar.org
• AmTrust, "Lawyers Professional Liability Application" (form LPLPRO-APP-01 0523, May 2023): amtrustfinancial.com
• AttPro (Berkshire Hathaway / MedPro), "What can lawyers gAIn from artificial intelligence?" (December 12, 2023): attorneyprotective.com
• FLMIC (Florida Lawyers Mutual), "AI Technologies Impacting Florida Lawyers and Their Clients" (October 3, 2023): flmic.com
• Berkley Insurance Company, "Artificial Intelligence Exclusion (Absolute)" form PC 51380 00 (06-24), endorsement PDF: Hunton-hosted carrier endorsement
• Hunton Andrews Kurth, "How Insurance Policies Are Adapting To AI Risk" (Law360, July 2, 2025): hunton.com
• Independent Agent Virtual University, "Verisk to Roll Out New General Liability Exclusions for Generative AI Exposures" (October 21, 2025): independentagent.com
• L Squared Insurance Agency, "Be Careful of New Artificial Intelligence Attorney Malpractice Insurance Exclusion!" (October 2024): l2insuranceagency.com
• LawSites, "Querious launches bar discount program with State Bar of Michigan as first partner" (October 2025): lawnext.com
• Oregon State Bar Professional Liability Fund, "Generative AI Policy Development Guidelines for Law Firms" (Rev. April 2026): assets.osbplf.org
• Oregon State Bar Professional Liability Fund, "Sample Generative AI Policy for Law Firms" (DOCX, January 2026): assets.osbplf.org
• OSB Formal Opinion No. 2025-205, "Artificial Intelligence Tools" (approved February 2025): osbar.org
• ORS 9.080(2)(a) (statutory authority for the PLF mandatory-coverage program): oregon.public.law
• Stan Sterna (Aon), quoted in "Insurers eyeing AI risk at CPA firms," Digital Insurance / Accounting Today (April 21, 2026): dig-in.com
• CNA Financial Corporation, FY2025 Annual Report on Form 10-K (accession 0000021175-26-000011, filed February 10, 2026): sec.gov
• The Hanover Insurance Group, FY2025 Annual Report on Form 10-K (accession 0001193125-26-060983, filed February 20, 2026): sec.gov
• NAIC Model Bulletin on the Use of Artificial Intelligence Systems by Insurers (adopted December 4, 2023): content.naic.org
• Colorado Division of Insurance Regulation 10-1-1 (2023), "Governance and Risk Management Framework Requirements for Life Insurers' Use of External Consumer Data and Information Sources, Algorithms, and Predictive Models." Cited via Colorado DOI. Primary URL not publicly stable.
• NYDFS Insurance Circular Letter No. 7 (2024), "Use of Artificial Intelligence Systems and External Consumer Data and Information Sources in Insurance Underwriting and Pricing": dfs.ny.gov
• Vanguard Specialty Insurance memo by Stephen Van Wert, "Artificial Intelligence Exclusions on Lawyers Professional Liability Policies" (October 2024): cited via specialty broker reporting; memo not publicly hosted at a stable URL.
• WTW Insurance Marketplace Realities 2026 (October 2025): wtwco.com
• Armilla AI / Chaucer (Lloyd's Syndicate 1084), Affirmative AI Liability Insurance launch (April 30, 2025): armilla.ai
• Armilla AI / Chaucer, "Vanguard AI" coordinated structure announcement (February 10, 2026): armilla.ai
• Counterpart, "Counterpart expands AI coverage for small businesses" (FinTech Global, November 26, 2025): fintech.global
• Lloyd's Market Association International Professional Indemnity Committee, "The Impact of Artificial Intelligence on the International E&O Market" (September 18, 2025): lmalloyds.com

Carrier and broker citations verified 2026-05-02. Public-filing references verified against EDGAR on the same date.

---