No AI Governance.
No Coverage.
That is where the specialty insurance market is heading, and most agents have never been asked to advise a client on it. Underwriters outside D&O and EPLI rarely think about it at all. Margaret Spence spent decades inside this industry before she built a framework to govern AI. She is not translating governance into insurance language. She already speaks it.
Of firms across the Lloyd's market now have, or are actively building, formal AI governance frameworks, up sharply from a year ago.
Source: Lloyd's Market Association AI Risk Management Survey, April 2026 (39 firms, 60%+ of market stamp capacity)Have adopted NAIC AI governance model bulletins. The EU AI Act's high-risk obligations, which explicitly cover insurance risk assessment and pricing, become enforceable August 2, 2026.
The Operational Guide For Your Clients
When AI Breaks the Law: AI Governance for Talent Leaders, by Margaret Spence
Your clients are being asked to prove AI governance to keep their coverage, and most of them do not know where to start. This is the book that walks them through it: what their exposure looks like, what documentation an underwriter will expect, and how to build a defensible program before a claim forces the conversation. Hand it to a client before renewal, use it to open a governance-readiness conversation on a new account, or build it into your agency's client education program.
Governance Is Becoming A Condition Of Coverage
This is not a regulatory mandate from a government body. This is the private market, actuaries, underwriters, and risk committees, independently arriving at the exposure regulators are also moving to police. When insurers and regulators agree on what constitutes dangerous AI, your clients have nowhere left to hide, and neither do you if you cannot advise them on it.
The Lloyd's Market Is Building Governance Alongside AI, Not After It
In April 2026, the Lloyd's Market Association surveyed 39 firms representing more than 60% of Lloyd's market stamp capacity. Ninety-three percent now have or are actively developing formal AI governance frameworks. Over 60% mandate human review of AI-generated outputs. Forty-four percent assign governance responsibility to the CTO, and 33% have stood up dedicated AI governance committees. LMA Head of Actuarial and Exposure Management Sanjiv Sharma put it plainly: "AI adoption across the Lloyd's market has accelerated quickly over the past 12 months, but what's encouraging is that governance is being built alongside it, rather than after the fact."
Underwriters Rated Professional Indemnity The Highest-Severity AI Loss Scenario
The LMA's companion survey gathered 144 responses, 94% of them underwriters, on the AI-related loss scenarios most likely to generate claims. Professional Indemnity, the coverage line for losses caused by errors, omissions, or negligent advice, came out on top. When AI is embedded in HR decisions, underwriting, claims handling, or client-facing advice and produces erroneous or discriminatory outputs, that is a PI exposure. Systematic algorithmic errors do not produce one claim. They produce thousands.
Standalone AI Liability Products Now Require Proof Of Governance To Bind
In 2024, Armilla AI became the first Lloyd's coverholder dedicated exclusively to AI liability, backed by Chaucer Group and Axis Capital. By January 2026, its Lloyd's-backed coverage expanded to $25 million as demand outpaced supply, largely because traditional insurers were retreating from the category entirely. To access that coverage, an organization must demonstrate ongoing AI model quality assessments, documented governance frameworks, performance monitoring, and vendor risk management. Testudo, a separate claims-made product launched in January 2026, ties coverage continuity directly to current documentation: let it lapse, and coverage lapses with it.
Standard Policies Are Being Rewritten To Exclude AI Entirely
The Insurance Services Office introduced two new CGL endorsements adopted across the market in early 2026: CG 40 47, which excludes all AI-related bodily injury, property damage, and advertising injury claims from standard coverage, and CG 40 48, a narrower version limited to advertising injury. At least six major carriers, including AIG, WR Berkley, Great American, Hamilton Insurance Group, and Philadelphia Indemnity, have filed their own AI exclusion endorsements. The result is a bifurcated market: clients who can demonstrate governance access affirmative coverage through specialist markets. Clients who cannot are left holding an expanding gap, right as AI-related litigation accelerates.
Sources: Lloyd's Market Association AI Risk Management Survey (April 2026); LMA AI Loss Scenarios Survey Results (January 2026); Armilla AI / Chaucer Group AI Liability Policy (2024–2026); Insurance Services Office CGL endorsements CG 40 47 & CG 40 48 (2026); EU AI Act Regulation (EU) 2024/1689.
What Underwriters Require Maps Almost Perfectly To What The EU AI Act Demands
| What Lloyd's Underwriters Require | EU AI Act Article |
|---|---|
| Documented AI governance frameworks | Article 17, Quality management systems |
| AI model performance monitoring | Article 9, Risk management systems |
| Human oversight mechanisms | Article 14, Human oversight measures |
| Third-party and vendor AI risk management | Article 28, Obligations of deployers |
| Bias and fairness testing documentation | Article 10, Data governance and management |
| Incident logging and reporting | Article 73, Reporting of serious incidents |
The Law Changed Last Week.
Did Your Client's Policy?
The AI Law Tracker monitors state, federal, and international AI legislation in real time, mapped to what it means for workforce decisions, vendor contracts, and governance gaps. Built for the advisor who cannot afford to be the last one to know before an underwriter asks.
Four Roles, One Blind Spot
Outside of specialty D&O and EPLI lines, most people in these roles have never been asked to think about AI governance. That gap is closing fast.
Retail Brokers
Fielding renewal conversations where a client's AI use is suddenly relevant to their coverage, and no framework for having that conversation.
Independent Agents
Advising commercial clients across hiring, HR, and operations tools without a way to flag where AI governance gaps are creating uninsurable exposure.
Underwriters
Outside of D&O and EPLI specialists, most underwriters have not yet built AI governance into how they evaluate risk. That is changing this year.
MGA Groups
Building underwriting appetite and client guidance for a risk category that barely existed two years ago and is now a stated market priority.
Practical Governance Readiness, Not A Compliance Lecture
You do not need to become an AI engineer. You need a working knowledge of what governance-ready looks like, and a way to bring that to your clients with confidence.
Client Governance-Readiness Workshops
Sessions your team can bring directly to commercial clients to walk through what an underwriter will expect to see.
Agency Team Training
Equip brokers, agents, and account managers to recognize AI governance exposure during renewal and new-business conversations.
SimpliFocus™ Governance Walkthroughs
A guided introduction to the 12-step governance system your clients can build a program around.
Client Education Toolkits
Playbooks and one-pagers your agency can put in front of clients without building the content from scratch.
Speaking & Conference Sessions
Keynotes and breakouts for agency conferences, MGA summits, and carrier producer meetings on the governance-coverage link.
Advisory Sessions
Direct consultation for underwriting and MGA teams building AI risk appetite and client guidance from the ground up.
Built On Insurance, Risk, And Underwriting Judgment
Margaret Spence understands the insurance sector from inside it. Her background includes work as an insurance adjuster, agent, insurance leader, and risk manager, combined with deep HR, workplace risk, and employment practices experience. $495 million in settled workplace claims. A $95 million single-client risk portfolio. She recently spoke for APCIA on Underwriting Insurance HR 2030. She is not translating AI governance into insurance language for the first time on this page. She has been bringing insurance judgment into governance work for years.
Also Worth A Look:
Let's Talk About Your Book Of Business
Tell us what you're seeing with clients, renewals, or underwriting appetite around AI. We'll help identify the right starting point, whether that's a workshop, a toolkit, or a conversation about copies of the book for your team.