IMA Financial Group has completed a $4 billion property insurance placement for a publicly traded artificial intelligence (AI) and high-performance computing (HPC) data center company.
Structured with a global property carrier, the deal is one of several billion-dollar programs the independent US brokerage has advised on in recent years, underscoring how AI-optimized campuses are reshaping demand for large-ticket property capacity.
IMA, which specializes in risk management, employee benefits and investment advisory services, began focusing on data center risk in the early 2000s. Its data center and digital infrastructure practice now serves commercial, hyperscale and digital asset operators across North America.
The placement comes amid a rapid global build-out of AI infrastructure. IDC estimated that AI-centric infrastructure spending reached about $90 billion in the fourth quarter of 2025 alone, with cumulative spend expected to exceed $900 billion by 2029.
Separate forecasts suggest total data center capex could more than double from around $430 billion in 2024 to $1.1 trillion by 2029 as AI transforms compute and power requirements.
Against that backdrop, AI and HPC campuses are emerging as a new stress test for insurers. Recent analysis has highlighted how AI data centers are creating capacity and credit risk challenges as short-lived, high-value chips are financed through long-dated debt structures and special purpose vehicles, with single locations sometimes carrying $10 billion–$20 billion in replacement cost.
Some market commentators have noted that, just a few years ago, it was “nearly impossible” to reasonably insure a $20 billion campus, but by 2026 such risks had become a regular topic of negotiation as more global and London market carriers committed dedicated line size and facility-style placements.
IMA said its latest deal reflects a long-running bet on digital infrastructure specialization.
"This placement reflects two decades of experience in the data center space, bringing that expertise to today’s most advanced facilities," said Patrick Datz, IMA digital risk practice leader. "AI and HPC data centers are among the most complex, capital-intensive assets in the world. They require an advisor who understands the technology and the risks, and can communicate both to a carrier market that is still getting up to speed."
As hyperscalers, colocation operators and AI-focused infrastructure firms deploy hundreds of billions of dollars in new construction and equipment, IMA argued that conventional commercial property programs are no longer sufficient.
“As investment in AI infrastructure accelerates, we’re seeing a growing need for insurance solutions that can scale alongside these assets,” said Rachel Nixon, IMA senior vice president and data center co-practice lead.
The IMA placement lands as insurers and brokers test new ways to support AI mega-projects.
A series of recent reports has highlighted how traditional property limits are being stretched by campuses valued in the tens of billions of dollars and financed with private equity and private credit.
To bridge the gap, carriers and intermediaries are increasingly looking to capital markets. Some insurers and brokers are using catastrophe bonds and other insurance-linked securities to bring in hedge funds and private investors to share losses on large data center developments, with lenders pressing for additional protection where conventional capacity is insufficient.
That points to a more layered risk-transfer stack for AI infrastructure, combining traditional property and business interruption cover with structured facilities, alternative capital and, in some cases, parametric triggers for events such as power outages or grid failures.
Policy wordings are also under pressure. Market commentary has flagged gray areas around inland storage of GPU inventory, the interaction between property, construction, cyber and equipment breakdown cover, and the way business interruption responds to AI training workloads that are highly time-sensitive and revenue-intensive.
IMA’s data center practice spans property, builders risk, cyber and technology liability, errors and omissions, directors and officers liability, crime and parametric business interruption. The firm said its advisory approach looks at how facilities are constructed, how they generate revenue and where traditional insurance structures fall short before a single submission reaches the market.
That approach reflects a broader trend of brokers and carriers building dedicated digital infrastructure teams that cut across property, cyber, technology E&O and D&O, as investors push into AI-heavy transactions and long-term power purchase agreements linked to data center growth.
As AI-related infrastructure spending heads toward the trillion-dollar mark over the next few years, broker-led facilities and alternative capital structures are likely to play a growing role in how the sector manages accumulation risk and supports the next wave of digital build-out.
