How the Insurance Market Is Responding to AI Risk
As you plan for 2027, which ranks higher on your priority list: artificial intelligence (AI) adoption and risk readiness, or outpacing competitors already using it well? For 38% of organizations, it’s nearly a tie, according to HUB International’s 2026 Middle Market Profitability & Resiliency Executive Survey.1 Yet interest in AI-related coverage dropped 11 percentage points year over year, from 46% to 35%, the sharpest decline of any risk category in the report.
That gap — more exposure paired with less appetite to insure it — is where the opportunity sits, even as AI insurance exclusions spread across commercial policies. Organizations that get precise about their AI exposure now can lock in stronger coverage terms while carriers are still building new products.
Since January 2026, the Insurance Services Office’s (ISO) generative AI exclusions have given carriers a standardized way to remove AI-related exposure from the coverage a commercial general liability policy provides.2 These generative AI exclusions in commercial general liability insurance are now in force at more than 2,300 filings across 49 states, filed by 60-plus property and casualty insurers.3
Where AI insurance coverage gaps show up
A British Columbia tribunal held a major airline responsible after its chatbot gave a customer incorrect bereavement-fare information.4 The tribunal rejected the airline’s argument that the bot was a separate, independent actor: The organization deploying an AI tool owns what that tool tells customers. A similar incident hit a software development tool provider whose AI support bot invented a login policy that led to customer cancellations.5 The lesson extends to any AI touchpoint with customers, employees or vendors, not just customer-facing chatbots.
New AI coverage is emerging
Carriers aren’t only pulling back. For example, Lloyd’s of London, through specialty insurer Armilla, now offers a form of AI liability insurance covering losses from malfunctioning AI tools, including chatbot hallucinations. Armilla has pointed to the airline example as exactly the kind of claim it would pay.6 The Financial Industry Regulatory Authority (FINRA) also added hallucination risk to its 2026 oversight expectations for AI governance.7 Specialist capacity is forming right where conventional coverage is retreating, and organizations that can clearly describe their AI exposure are best positioned to access it on favorable terms.
Questions to bring to your next advisor conversation
- Does your privacy liability coverage reflect how AI systems collect and process data?
- What AI exclusions were added at your last renewal, and what do they really remove?
- If an AI tool causes a loss, who’s on the hook: you, your vendor or both?
- When was your coverage last reviewed against this year’s policy language?
- Do you know everywhere AI is running in your business, including AI your vendors turned on without a specific procurement event?
New AI insurance exclusions and tighter underwriting are already appearing at renewal. Bring these questions to your HUB advisor before that renewal lands, not after.