Counterparty
Identify the actual obligor, parent support, tenant credit, and any non-recourse project entity between the contractor and the ultimate economic user.
The finance shift
The opportunity is moving from cash-funded technology spend into leveraged infrastructure.
For years, AI infrastructure could be viewed primarily as a technology-company capital-expenditure story. The next phase is broader. A recent Brookings analysis estimates U.S. AI infrastructure investment could total about $10.3 trillion between 2025 and 2032. Morgan Stanley estimates roughly $2.9 trillion of global data-center capex through 2028, with approximately $1.5 trillion requiring external financing after hyperscaler cash flows.
That changes the management question. The technology thesis may remain compelling while the financing structure becomes progressively more traditional: project companies, leases, bank debt, corporate bonds, private credit, securitization, guarantees, completion risk, and collateral.
$1.4TEstimated hyperscaler cash-flow funding
$1.5TEstimated financing required from outside capital
52%Approximate share of the $2.9T spend requiring external capital
The capital stack
What builders should underwrite
A hyperscaler logo somewhere in the deal is not the same thing as payment security.
For contractors and infrastructure operators, the most important questions are often one or two layers below the headline tenant. Who is the contractual counterparty? Who funds the project company? What is guaranteed, by whom, and under what conditions? Which party absorbs schedule delay, escalation, power delay, commissioning failure, or scope growth?
The same discipline applies to backlog quality. A large data-center award can be strategically attractive and still create working-capital stress if mobilization, equipment, labor, or subcontractor cash requirements arrive well ahead of billing and collection.
Completion
Understand who owns schedule risk, commissioning risk, permitting delay, liquidated damages, and cost overruns.
Power
Confirm energy availability, utility timing, interconnection dependencies, and which party carries the cost of delay.
Cash conversion
Model deposits, procurement, mobilization, billing milestones, retainage, collections, and the borrowing-base impact before treating backlog as liquidity.
What lenders should underwrite
The demand thesis does not eliminate structure risk.
AI demand can remain strong while individual credits underperform. The lender still has to underwrite lease terms, tenant concentration, collateral value, takeout assumptions, construction completion, refinance risk, utilization, and the durability of the underlying economics.
Recent SoftBank high-yield issuance tied in part to OpenAI investment and the market scrutiny around Oracle-linked Project Jupiter illustrate the migration of AI risk into broader credit markets. Neither example proves a systemic problem. They do show that leverage and execution are becoming inseparable from the technology narrative.
What CFOs should watch
The balance sheet has to keep pace with the backlog.
For companies participating in the buildout, the finance function should translate growth into explicit liquidity and risk requirements. That means scenario modeling around project starts, procurement, bonding, equipment, labor ramp, receivables, customer concentration, lender capacity, covenant headroom, and downside cases.
The biggest operational danger is confusing revenue opportunity with funded capacity. A company can win extraordinary work and still outrun its cash, surety, or borrowing base.
AI may be new. Working capital, leverage, and counterparty risk are not.
The management takeaway
Opportunity and underwriting have to scale together.
The AI infrastructure cycle may create one of the largest construction and financing opportunities of the decade. The right response is not to become less enthusiastic. It is to become more disciplined about who is financing the project, where authority sits, where loss can land, and whether the business has the balance-sheet capacity to participate safely.
Owner and CEO conversation
Pressure-test the opportunity before the capital is committed.
KSIG can help construction and infrastructure companies model project cash, lender capacity, working-capital requirements, equipment financing, bonding, counterparty exposure, and downside scenarios before growth becomes a balance-sheet problem.
Discuss the opportunity