Premium impact
Specialty brokers report 5–15% premium loadings on policies where AI agent exposure is undocumented, and full exclusions where the insured cannot evidence operational logging.
Lloyd's syndicates are pricing AI exposure into cyber and Tech E&O policies. The submission that shows working AI agent logging — with a verifiable hash chain and a regulator-recognised pack format — meaningfully changes underwriter loadings.
Following Air Canada's chatbot ruling and the wave of incident disclosures through 2025, the cyber and E&O markets at Lloyd's have moved from "no specific AI exclusion" to "show your AI logging methodology at submission."
Specialty brokers report 5–15% premium loadings on policies where AI agent exposure is undocumented, and full exclusions where the insured cannot evidence operational logging.
Major syndicate submissions now include AI-specific questions: which models in production, governance framework, incident response procedure, and audit logging methodology.
When a claim is made under a cyber or E&O policy involving an AI-driven service, the insurer's investigators will request a reconstruction of the relevant agent actions. Without tamper-evident logs, the claim is harder to defend.
The Agent Audit insurance evidence pack is purpose-built for submission at renewal or for claim support. It demonstrates the controls underwriters want to see, in the format they're used to receiving.
Composite questions across Lloyd's specialty cyber and Tech E&O submissions in 2025–2026: which agents you run, what classification of data they touch, how you log and retain action records, whether the logs are tamper-evident, and how you would prove an incident timeline after the fact. Our renewal guide walks through what counts as a defensible answer to each.
It demonstrably affects loadings. Underwriters price AI exposure by assessing whether your control environment can produce contemporaneous, tamper-evident evidence of agent behaviour. A working audit trail with regulator-recognised pack formats is treated favourably by every Lloyd's syndicate we have spoken to.
The Insurance Claim pack bundles incident timeline, affected receipts, integrity proof, RFC 3161 notarisation tokens, redacted inputs and outputs, and per-action attribution into a single signed artefact brokers and loss adjusters can submit alongside the claim notification.
We're broker-agnostic. The pack is portable JSON + PDF — your broker can submit it alongside your renewal questionnaire regardless of which house they sit in. No broker-specific lock-in, no integration required on their side.
The market is emerging. Tech E&O and specialty cyber markets are starting to address AI-specific exposures, and underwriters are asking how you would evidence agent behaviour after an incident. Tamper-evident logging is increasingly a prerequisite for the AI-related questions in submission questionnaires.
Hand your broker the AI risk evidence underwriters are starting to ask for. The pack is portable, signed, and ready to drop into your renewal submission.