Corgi expands into AI data center insurance as infrastructure risks and GPU values climb

Corgi has launched a dedicated data center insurance program covering risks from construction and commissioning through operational property, equipment, business interruption and technology-related exposures, as the artificial intelligence infrastructure boom creates a rapidly expanding market for specialist insurance. The San Francisco-based insurance technology company is targeting colocation operators, GPU cloud and AI compute providers, enterprise and edge data centers, and developers constructing new facilities. Rather than treating a data center primarily as commercial real estate, Corgi is structuring coverage around the interconnected risks created by power systems, cooling infrastructure, computing equipment, construction schedules and service availability. The opportunity is significant, but the commercial test will be whether Corgi can translate that broader risk architecture into competitive underwriting capacity, usable policy terms and claims performance as the value concentrated inside AI facilities continues to rise.

Corgi announced the program on October 2, 2026, describing it as an insurance framework covering the life of a data center from initial construction through continuing operations. During development, customers can review builders risk and delay-in-start-up protection for covered physical losses that postpone scheduled operations, while operating facilities can consider property, equipment breakdown, business income, cyber and liability exposures. Corgi is also exploring residual-value protection around graphics processing units, potentially addressing a different problem from conventional property insurance by helping support the future value assumptions attached to expensive computing hardware. The company’s current product information makes clear, however, that data center insurance can involve several policies or coverage sections rather than a single universal policy, with availability, limits, deductibles and terms dependent on the individual risk.

Why is the AI data center boom creating a much larger specialist insurance opportunity?

The timing of Corgi’s launch coincides with one of the largest infrastructure investment cycles generated by artificial intelligence. Data centers increasingly contain not only valuable buildings and conventional electrical equipment, but dense clusters of high-performance processors, sophisticated cooling systems, backup power infrastructure and networking equipment whose failure can interrupt revenue-generating computing workloads. As facilities become larger and more capital intensive, the financial consequences of construction delays, equipment failure, fire, natural catastrophes or extended outages can increase sharply.

Allianz Commercial estimated in August that annual data center investment could rise from roughly $500 billion in 2024 to more than $1 trillion as early as 2027. The insurer projected that the global data center insurance market could increase from around $11 billion to more than $24 billion by 2030 as capacity expands, insured values rise and operators require broader combinations of construction, property, engineering, business interruption, cyber and liability protection. Allianz also noted that the construction cost of a single large AI campus can exceed $20 billion, before considering the substantial value of high-performance computing hardware installed inside the facility.

Those economics change what insurers are being asked to protect. A traditional property policy may address physical assets, but developers and operators can also face lost income when a covered event delays commissioning, equipment failures that disrupt computing capacity, utility interruptions, cyber events and contractual liabilities to customers expecting defined service levels. Corgi’s decision to organize the program around multiple stages of the facility lifecycle therefore reflects an insurance market that is following data centers as they evolve from specialized buildings into critical digital infrastructure.

What exactly does Corgi’s data center insurance program cover across construction and operations?

Corgi’s program begins before a data center enters service. Its construction and go-live offering includes builders risk for work in progress and potential delay-in-start-up coverage where covered physical damage pushes back the planned opening date. For developers financing large facilities against an expected commencement of commercial operations, a significant delay can create economic consequences beyond the cost of repairing the damaged construction itself.

Once operational, the risk profile changes. Corgi’s program allows customers to review coverage for buildings, servers, power equipment, cooling infrastructure and related systems, together with equipment breakdown protection and business income or extra-expense coverage after certain covered interruptions. Power and cooling are particularly important because computing hardware depends on continuous electricity and tightly controlled operating temperatures, meaning failures elsewhere in the facility can affect hardware that was not itself the initial source of the incident.

The program also incorporates cyber and service liability considerations rather than assuming every technology-related loss belongs inside conventional property insurance. Corgi specifically advises customers to examine the boundaries among cyber, property, technology errors and omissions and business-interruption policies because the trigger for a loss can determine which policy responds. A cyberattack that causes an outage, for example, may involve different insurance questions from mechanical equipment failure, even if both events ultimately remove computing capacity from service.

Corgi is consequently selling coordination as much as an individual insurance product. The company’s own guidance states that data center insurance is not necessarily one policy and that the appropriate combination depends on the facility, contracts, equipment ownership and stage of development. That qualification is commercially important because the value of the program will depend on how effectively different coverage sections work together when an incident crosses conventional insurance boundaries.

Why could GPU residual-value protection become important to financing AI infrastructure?

The most distinctive element in Corgi’s announcement concerns GPUs. Modern AI data centers can concentrate enormous amounts of capital in processors whose economic value depends not only on whether the equipment continues functioning, but also on how quickly newer generations of hardware change the market value of existing machines. That creates a financing question that ordinary property insurance does not necessarily address.

Residual-value protection focuses on what an asset may be worth at a future point rather than merely compensating for physical loss or damage. In principle, greater certainty around the future value of GPU equipment could make the hardware easier to finance or support financing structures in which lenders and investors depend partly on the equipment’s recoverable value. This issue becomes increasingly relevant as smaller AI cloud operators and infrastructure companies seek capital to acquire expensive processor fleets without carrying all of the financing risk themselves.

Corgi’s October announcement said the company is introducing residual-value protection intended to help protect GPU hardware as a financeable asset. Its more detailed current product page is appropriately narrower, telling prospective customers to ask whether a separate residual-value arrangement is available and what terms would apply. The distinction means GPU residual-value protection should not be interpreted as an automatic component of every Corgi data center insurance package.

That product area could nevertheless become strategically important. Insurance that improves the financeability of computing assets would place carriers closer to the capital structure of the AI infrastructure boom rather than limiting their role to reimbursing physical losses after an incident. The opportunity also carries underwriting difficulty because GPU values depend on technology cycles, secondary-market demand, useful economic life and the arrival of newer processors, requiring assumptions that differ substantially from conventional property-loss modeling.

How do fire, cooling failures and natural catastrophes change data center underwriting?

Physical resilience remains fundamental despite the technology-heavy nature of the industry. Allianz Commercial found that fire was the largest contributor to the severity of data center losses in the claims information it analyzed, accounting for well over half of approximately €700 million in losses, while water damage was the most frequent source of claims. Power failure and natural catastrophe exposures add further complexity because a facility can lose operating capacity without damage originating directly inside computing equipment.

Location also increasingly matters as data center construction moves into markets constrained by power availability, water resources and land. Allianz estimated that about 79% of global data center capacity is located in areas with heightened exposure to natural catastrophes, while more than half is exposed to chronic heat and drought stress. Those risks can interact with cooling and energy requirements, meaning the underwriting process has to consider not only the design of the building but the resilience of the surrounding infrastructure on which the facility depends.

This creates a natural use case for more granular underwriting. Two facilities with similar computing capacity can have materially different risk profiles depending on grid redundancy, cooling architecture, fire suppression, flood exposure, backup generation, equipment concentration, security controls and customer contracts. Corgi asks prospective customers to provide information on those characteristics alongside equipment inventories, ownership arrangements, revenue dependencies, security measures and loss history.

For an AI-native insurer, the potential advantage is using technology to ingest and evaluate that information more quickly while maintaining sufficient underwriting discipline. Faster quoting alone will have limited value on complex infrastructure if the underlying risk has not been priced correctly. Data centers therefore offer Corgi an opportunity to demonstrate whether its technology-led model can move beyond relatively standardized commercial insurance into risks involving substantial engineering and operational complexity.

How does the new data center program fit Corgi’s rapid expansion across commercial insurance?

The launch is part of a much wider expansion by Corgi since the company received regulatory approval in 2025. Founded by Emily Yuan and Nico Laqua, Corgi has been building what it describes as a full-stack, AI-native insurance platform, combining underwriting technology with policy administration, claims operations and expansion into new commercial insurance categories. Its product portfolio has moved beyond startup insurance into additional areas including artificial-intelligence risk, trucking and other specialist exposures.

That expansion has been supported by substantial private capital. In May 2026, Corgi announced a $160 million Series B led by TCV at a $1.3 billion valuation, following $108 million of previously disclosed seed and Series A financing. Only three weeks later, the company announced another $106 million Series B1 led by TCV, saying that transaction valued the business at $2.6 billion.

Corgi’s own currently published financing pages therefore document an unusually rapid sequence of capital formation, giving the company resources to enter insurance markets that require specialist underwriting teams, regulatory infrastructure and claims capabilities. The data center program is strategically different from simply selling another software tool because insurance expansion requires actual risk-bearing or carrier relationships, regulatory compliance and sufficient capacity behind the policies.

The legal and underwriting structure is also important. Corgi Insurance Services, Inc. states that it acts as a licensed insurance producer and program administrator rather than the insurer itself. Coverage may be underwritten through affiliated or partner carriers, including Corgi Insurance Company, Inc., an admitted property and casualty insurer, while certain coverages may also involve Technology Risk Retention Group, Inc. or other partner carriers depending on jurisdiction, product and risk characteristics.

Can Corgi differentiate when established insurers are also chasing data center growth?

Corgi is entering an attractive market, but it is not entering an empty one. Major commercial insurers and brokers already serve data center operators, infrastructure developers and technology companies, and the projected expansion of insured values is encouraging established market participants to develop increasingly sophisticated products around construction, cyber, operational continuity and equipment risk. Corgi therefore needs differentiation beyond the existence of a dedicated data center program.

Its strongest potential advantage is the ability to design underwriting infrastructure around newer operating models instead of adapting legacy products incrementally. GPU cloud providers, AI compute hosts and rapidly expanding infrastructure companies may have risk profiles that do not fit neatly into assumptions built around conventional commercial property. Combining digital underwriting with coordinated construction, operational, technology and hardware-value coverage could make Corgi attractive to customers seeking fewer gaps among multiple insurance relationships.

The counterweight is underwriting history. Established insurers have decades of engineering data, catastrophe modeling, claims experience and relationships with large corporate risk managers, while infrastructure customers and lenders care about financial strength and dependable claims payment as much as user experience. Corgi’s program will need to demonstrate that technological speed can coexist with conservative risk selection and adequate capacity.

Pricing will provide another test. AI infrastructure is attracting enormous capital, but that does not make every risk attractive to insure. Concentrated GPU values, complex power arrangements, supply-chain constraints, extreme weather exposure and dependence on high facility availability could produce expensive claims, so sustainable growth depends on premiums and policy terms accurately reflecting those exposures rather than using aggressive pricing merely to acquire market share.

Why could data center insurance become part of the financing architecture for AI growth?

The broader opportunity extends beyond protecting buildings after they are completed. Large infrastructure projects typically require insurance throughout construction because lenders and investors want protection against events capable of delaying completion or impairing assets securing the financing. As AI campuses become larger and hardware values climb, insurance can become an increasingly important component of whether capital providers are comfortable funding development.

GPU residual-value protection takes that connection further. If insurers can structure credible protection around the future value of computing equipment, insurance could help distribute some of the asset-value risk that would otherwise remain with lenders, lessors or infrastructure operators. That would make the sector strategically relevant to financing structures supporting the AI build-out, although the viability of individual products will depend on underwriting terms and the ability to assess rapidly changing hardware economics.

Corgi is positioning itself at precisely that intersection of technology, infrastructure and insurance. The company does not disclose premium expectations, underwriting capacity, customer commitments or policy limits for the new program, so it is too early to determine how much business the launch will generate. Those figures will ultimately matter more than the breadth of the initial product announcement.

The next proof point is therefore commercial adoption accompanied by evidence of disciplined underwriting. Winning data center developers, colocation operators and AI compute providers would demonstrate demand, while successful claims handling and sustained carrier capacity would show whether Corgi can manage the complexity it is taking on. In a market where billions of dollars of infrastructure can depend on uninterrupted power, cooling and computing, insurance credibility will be earned when something goes wrong rather than when the policy is first sold.

What are the key takeaways from Corgi’s data center insurance program launch?

• Corgi launched a dedicated data center insurance program on October 2, 2026 covering risks from construction through ongoing operations.
• The target market includes data center developers, colocation companies, GPU cloud and AI compute hosts, enterprise operators and edge facilities.
• Construction coverage can include builders risk and delay-in-start-up protection following covered physical losses.
• Operating exposures can include property, servers, power infrastructure, cooling equipment, business interruption, cyber and liability risks.
• Corgi’s data center offering can involve multiple policies or coverage sections rather than one universal insurance policy.
• The company is exploring GPU residual-value protection, although availability and terms may require a separate arrangement.
• Allianz Commercial projects the global data center insurance market could grow from roughly $11 billion to more than $24 billion by 2030.
• Data center underwriting is becoming more complex as facility values, GPU density, power requirements and natural catastrophe exposures increase.
• Corgi’s own disclosures identify Corgi Insurance Services, Inc. as a program administrator, with coverage potentially provided through affiliated or partner carriers.
• Commercial adoption, underwriting capacity, pricing discipline and claims performance will determine whether Corgi can convert AI infrastructure growth into a defensible insurance business.