Underwriting the Unknown: Insurability as a Deep-Tech Diligence Signal

UK deep-tech founders are accustomed to technical milestones: a working prototype, a first paid pilot, a patent granted. Then a broker asks for the product liability proposal form and the conversation stalls. The material is novel, the failure modes are not in the actuarial tables, and the customer contract requires cover that no standard policy will provide. The company has not run out of engineering ambition. It has run into the insurance desert.

The gap is specialist, not absolute

Britain has one of the world's deepest insurance markets. Lloyd's syndicates, specialist managing general agents and brokers in London can write bespoke risks that would baffle a high-street insurer. The problem is not a lack of postcodes. It is a lack of capacity and expertise in the specific risk classes that deep tech creates: product liability for a first-of-a-kind material, professional indemnity for an AI-enabled inspection service, export risk for dual-use photonics, or property and business interruption cover for a quantum facility with unusual cryogenic and electrical hazards.

For a spinout clustered around Cambridge, Bristol, Oxford or Harwell, the nearest specialist broker may be in London and the nearest underwriter with relevant technical literacy may be in another market entirely. Catapult centres, Innovate UK and university technology-transfer offices can help founders prepare evidence, but they do not underwrite. The British Business Bank can support finance, and UK Export Finance can support export transactions, but neither substitutes for a commercial insurance programme. The desert is a capacity and expertise gap, not a simple absence of providers.

How underwriters price technical uncertainty

Insurers are not refusing to engage because they dislike innovation. They are translating uncertainty into contract terms. Where historical claims data is sparse, underwriters ask for failure modes and effects analysis, safety cases, test evidence, quality management systems, supply-chain traceability and a clear incident-response protocol. They look at customer concentration, contractual indemnities, recall exposure and the company's ability to survive a product failure. Where the evidence is thin, they narrow cover, add exclusions, raise deductibles, impose warranties or decline outright.

That process is a form of technical due diligence conducted by a different profession. It can be uncomfortable for founders, but it is informative. An insurer that has reviewed the evidence and agreed to write cover is offering a third-party view on whether the risk is bounded and transferable. An insurer that declines is not necessarily right, but the reasons for decline belong in the investment memo.

Insurability belongs in diligence

Investors should treat insurability as a diligence signal rather than an afterthought. The questions are straightforward. What cover exists today? What is excluded? Which risks are contractually allocated to customers or suppliers? Does a public procurement or export contract require insurance that the company cannot obtain? Is insurance budgeted in the runway plan? Does the board have adequate directors' and officers' cover, and does the company understand its product liability and professional indemnity exposure?

The answers affect valuation, deal terms and post-investment support. A company that cannot explain its insurability may also struggle to sell into regulated markets, pass customer procurement or scale beyond bespoke pilot agreements. Insurability is not binary. A startup may begin with limited cover, contractual risk allocation and a clear path to a broader programme. What matters is whether the path is credible and funded.

Practical steps for founders in 2026

First, start early. Begin conversations with a specialist broker twelve to eighteen months before the first commercial shipment, not after a customer demands cover. Second, build an evidence pack that an underwriter can use: failure modes and effects analysis, safety case, test logs, quality management documentation, supply-chain maps and export classifications. Third, stage the cover. A prototype or pilot programme may need a narrower policy than a scaled commercial deployment; do not assume full cover will be available on day one.

Fourth, use contractual risk allocation deliberately. Indemnities, liability caps, warranties and customer insurance requirements are not substitutes for insurance, but they can make a risk financeable. Fifth, engage regulators and sector bodies early where the product touches medicines, aviation, nuclear or transport. Sixth, budget insurance as a percentage of revenue and include it in pricing. Seventh, give a board-level owner responsibility for insurance and report progress to investors.

Where cover is unavailable, document why and what mitigations exist. That record is valuable in the next funding round and in customer negotiations. The insurance desert will not be solved by a single policy product. It will be mapped by founders who treat cover as a design constraint and by investors who ask about it before the cheque clears.