Arbol Introduces Coverage for Data Center Construction Moratoriums

Local resistance is creating a distinct financing risk for US data center projects: construction can be halted by government action even when a site suffers no physical damage. For developers, that creates a gap between having a viable site and being able to build on it. Arbol, a New York-based boutique insurer, has introduced coverage for developers and financiers facing that exposure, offering a way to transfer some of the risk of a prolonged government-imposed pause.

The issue matters because data centers require substantial construction commitments, while local approvals can change as projects advance. A delay may disrupt schedules and financing plans without damaging the property itself. That distinction is central to the proposed coverage: it responds to a political or administrative event that conventional property insurance may not treat as a covered loss.

A coverage gap amid project delays

Conventional commercial policies generally require physical property damage to trigger payment, while political risk insurance has typically focused on sovereign expropriation abroad. Local decisions that suspend construction—such as votes by county boards—can therefore sit outside standard coverage. The gap is not simply that a project faces opposition; it’s that the resulting interruption may fall between insurance categories designed for other kinds of losses.

Robert Hartwig, director of the Risk and Uncertainty Management Center at the University of South Carolina, said the product could ease concerns among investors in AI infrastructure. A policy that pays after a defined government action could give some investors a clearer account of one risk when assessing a project. But the terms are tailored, and local risks vary, so the product may be difficult to scale quickly across jurisdictions.

The market pressure is visible in project data. Data Center Watch, a monitor operated by AI intelligence firm 10a Labs, reported that at least 106 US data center projects, representing about $198 billion in investment, were blocked or delayed amid local opposition during the first half of the year. Those figures combine blocked and delayed projects, and they describe investment represented by the projects rather than a direct measure of losses. Even with that limitation, the scale suggests that local opposition has become a material planning issue for the sector.

Resistance has spread as the US midterm elections approach. Lawmakers in states including Texas and New York have enacted temporary pauses on new development, while in Pennsylvania, Gov. Josh Shapiro signed an executive order in August tightening oversight of data center projects despite having previously supported AI investment. These examples show how political support for a broad industry goal can coexist with closer scrutiny of specific projects. For financiers, that can make the path from policy support to local permission difficult to read.

Pricing depends on local conditions

Arbol says the policy is available nationwide, but the price varies with location, permit category and project exposure. Nationwide availability doesn’t mean every project can obtain coverage at a workable price: the insurer excludes areas with active moratoriums, which Jha said would be prohibitively expensive to cover. That boundary also points to a basic underwriting problem. When a pause is already in force, the risk has shifted from uncertain possibility to an existing condition.

Underwriting draws on local measures including water availability, grid capacity, electricity prices and the jurisdiction’s track record with data center development. Those factors can help describe a project’s operating context, but they don’t make future public decisions predictable. Similar systems typically require insurers to distinguish between risks that can be priced from available information and events that remain difficult to estimate. Here, the local political response is part of the exposure being assessed, rather than a background detail.

(And this is the part most people overlook.) A policy’s value depends not only on whether a covered event occurs, but also on whether its definition, waiting period and payout fit the project’s actual financing needs. The source terms are described as tailored, so prospective buyers will need to understand how those conditions apply to their own plans.

“It’s not that dissimilar to what we have done in climate risks,” Jha said. He added that Arbol has remained profitable while covering extreme weather and climate hazards that many mainstream insurers consider difficult to insure. The comparison explains the company’s approach: assess exposures that can vary by location and design coverage around them. Regulatory decisions, however, involve a different set of uncertainties from weather hazards, and experience in one category doesn’t by itself establish how the new product will perform.

Growth claims and the test ahead

Arbol recorded roughly $225 million in gross written premiums last year, compared with $2 million in 2020, when it began offering insurance. Jha said the firm was on track to double last year’s figure in 2026. That growth provides context for the company’s capacity to expand its business, but it doesn’t show how much of its premiums come from this new policy or how claims on the coverage will develop. The numbers here are underwhelming as evidence of this product’s track record: the policy is new, and the source offers no claims history for it.

Jha said companies representing tens of billions of dollars in construction value had lined up to buy policies since broker outreach began in September. That signals interest from project participants, though construction value isn’t the same as premiums written, coverage limits or likely payouts. If you’re working in this space, that distinction matters when judging whether early demand reflects a durable insurance market or an initial response to a visible financing concern.

What the policy could mean for projects

The policy’s performance will depend on how insurers assess changing local politics and how developers respond to community concerns. Jha said hyperscalers are refining their engagement with local communities and noted that many data centers still get built. Insurance may help address a defined portion of the financial risk, but the source doesn’t suggest it can settle local disputes or guarantee that a delayed project will proceed.

For developers and financiers, the broader implication is that approval risk may need to be considered alongside familiar questions about site conditions, power and water access. Coverage could make some projects easier to evaluate if its triggers and payout terms align with the exposure. Yet a policy that excludes places with active moratoriums will be least available where the risk has already become explicit. A useful risk transfer tool, with limits.

That said, the product’s existence is itself a sign that local opposition is being treated as a financing issue with potential insurance value. Whether that translates into a larger market will depend on pricing, the clarity of covered events and buyers’ experience with claims. Many projects still get built, as Jha points out, but that doesn’t make the delays immaterial to the companies funding them.