Mayflower and Hadron launch AI liability program

Per a Business Wire press release dated June 24, 2026, Mayflower Specialty, Ltd. (a New York managing general agent) and Hadron (a multinational specialty carrier) launched what they describe as the first dedicated affirmative artificial intelligence liability program in the United States, issued by Hadron and underwritten by Mayflower. The policy suite provides explicit coverage across the management-liability triad - directors and officers (D&O), employment practices liability (EPL), and errors and omissions (E&O) - for enterprise companies deploying AI, corroborated by Coverager. The product combines affirmative grants for AI-related management liability, employment practices, and professional liability with a difference-in-conditions and excess layer designed to drop down where legacy policies are silent, sub-limited, or exclusionary, per Business Wire and Reinsurance News.

Per the press materials, Mayflower underwrites AI-specific risks including model bias, model drift, and hallucinations using an auditable scoring engine aligned to NIST and ISO standards. The launch materials also cite McKinsey's 2025 State of AI report that 88% of organizations have integrated AI into at least one business area, using that statistic to frame market need. Coverager reports the program is backed by institutional reinsurance partners with placement structured by Aon Reinsurance Solutions, and notes Hadron has more than $250 million in committed capital from Altamont Capital Partners and other investors.

The product's stated appetite for bias, drift, and hallucination suggests underwriters are mapping traditional professional-liability constructs onto model-specific failure modes. That creates two practical implications: underwriters will need evidence - auditable logs, evaluation metrics, and governance artifacts - to rate exposures, and a scoring engine aligned to NIST/ISO language implies a checklistable baseline for controls that vendors and internal ML teams can target when seeking coverage. These are industry-pattern observations, not claims about either firm's internal underwriting thresholds.

Industry reporting frames this as a market-first US offering (Business Wire; Reinsurance News; Coverager). Similar affirmative-coverage launches elsewhere have tended to start narrow and priced conservatively, expanding as loss history and precedents accumulate - meaning risk managers who engage early should expect limited initial capacity, strict conditions, and tight claims definitions.

Documenting evaluation datasets, drift-monitoring telemetry, model-change logs, and governance decisions creates the artifacts underwriters will likely require. This is general industry advice, not a claim about Mayflower's specific internal requirements. If affirmative AI liability products scale, they could shift the calculus of acceptable residual risk for boards and CISOs by making some governance failures financially transferable and others explicitly uninsurable based on contract language.

Jeremy Epstein, Founder and CEO of Mayflower, said: "Every major technology wave has created a new class of liability, and the underwriters who showed up early with genuine standards defined that market for a generation." Pete Buccola, Group Chief Underwriting Officer of Hadron, is quoted describing the need to both limit exposure and consider underwriting approaches for consumers.

Three indicators matter going forward: the wording and claims definitions in issued policies (whether terms like "hallucination" and "bias" become operationalized, measurable trigger events); whether the auditable scoring engine publishes benchmarks or stays proprietary; and whether reinsurance capacity and placement dynamics expand, which would broaden pricing and limits. Whether this product materially shifts litigation or regulatory outcomes depends on claim frequency, severity, and how courts interpret policy language - public reporting does not yet provide loss experience or detailed policy wording beyond the high-level structure.