David Solomon, CEO of Goldman Sachs, says it remains uncertain how demand for capital and computing power will evolve as the AI industry matures. Jane Fraser, CEO of Citi, meanwhile, believes investment across the AI ecosystem is ongoing and there is still not enough compute capacity.

According to a Reuters report dated October 14, 2026, Goldman Sachs CEO David Solomon stated that it remains uncertain how demand for capital and computing power will evolve as the AI industry develops. Citi CEO Jane Fraser, on the other hand, holds that investment across the AI ecosystem is continuing and compute supply is still insufficient.

For international students targeting finance roles in the U.S., news like this can serve as material for industry research and interview preparation. After reading about “rising AI financing,” ask a follow-up question: Who receives this capital, how will they pay it back, and under what scenarios might repayment fail? This gives your financial analysis a concrete focus.

01 The AI construction boom raises familiar finance questions

Building a data center requires land, power, hardware and construction funding. Corporate cash reserves often cannot cover the full outlay, so banks step in with loans, bond issuances and related financing structures.

A Reuters background report from June 29 explained that tech firms raise capital via bonds denominated in multiple currencies, while banks structure financings around pre-signed data center leases. Future revenue locked in by these leases forms a core basis for evaluating such transactions.

Finance students familiar with corporate finance can readily connect the dots: Why is the company borrowing? Does the financing term align with project payback timelines? Will cash flow suffice to cover interest payments? These points fit naturally into capital structure and cash-flow analysis.

You do not need to become an expert on foundation models to research AI-linked finance. You should, however, understand the relationships among developers, compute tenants and lenders, and know where revenue originates and where costs accrue. Otherwise, even if you memorize financing figures, you will struggle to discuss the underlying transaction.

02 Beyond the growth narrative, banks focus on debt service capacity

On October 14, Reuters reported that rising rate expectations have amplified market concerns over AI infrastructure financing, dragging down share prices for several exposed firms. For companies borrowing to build data centers and purchase chips, higher funding costs compress returns.

Two days later, Reuters covered mounting debt pressure at CoreWeave. Some AI financings are backed by GPU assets and long-term customer contracts. Lenders must assess both asset values and whether contracts will reliably generate ongoing revenue.

This does not directly signal a reversal in AI demand. A hot industry for a firm and that firm’s ability to service debt are separate assessments.

Long-term contracts appear to stabilize revenue, but when customers are concentrated among a small group of counterparties, deeper analysis is required: Can contract revenue cover operating expenses? If the project is delayed and revenue arrives later, how will interest be serviced? If customer demand shifts, do the original cash-flow projections still hold?

Growth rates, profit margins and discount rates then gain grounding in business realities. When building models, you must justify your assumptions and identify which changes would invalidate your conclusions.

03 Different teams within the same AI firm track different metrics

This is highly relevant for finance students preparing for recruitment. Analysis tailored to your target role lands better than a broad overview of the AI sector.

If targeting investment banking or debt capital markets, focus on financing choices. Why is the company raising capital now? Why loans versus bonds? Do tenors match the use of proceeds? If you can only describe the hype around the business without explaining the transaction structure, your research is incomplete.

Credit research and private credit prioritize repayment. When does debt mature? Which customers drive cash flow? What alternate liquidity buffers exist if revenue undershoots? When reading coverage on CoreWeave, bring these questions to public filings for supporting evidence.

Equity research and asset management focus elsewhere. How much revenue growth will follow higher capital expenditure? How will utilization, pricing and depreciation shape profitability? Even with a strong business outlook, you need to gauge how optimistic the current valuation is.

Risk management candidates can build scenario analyses around interest rates, project delays and demand shocks. Identifying the most sensitive variables and cash-stress triggers is more valuable than delivering a single point forecast.

04 Research projects on your resume must withstand follow-up questions

When relevant internships are missing, independent research can demonstrate analytical skill — but it should never be framed as hands-on transaction experience. Pick a publicly transparent AI infrastructure company and study whether its cash flow supports expansion; there is plenty to unpack.

Cite sources for data pulled from annual reports, quarterly filings and investor decks. Clearly flag your own assumptions for growth, interest rates or delays. Leave gaps where information is unavailable instead of inventing numbers to polish the report.

Tweak one or two key variables in Excel and observe impacts on cash balances, interest coverage or valuation. The model does not need to be complex, but units, formulas and outputs must be auditable. Condense findings into a one-page brief stating your thesis, supporting evidence and top outstanding risks.

On your resume, document the research subject, methodology and tangible deliverables. This gives you concrete talking points in interviews: Why this variable? Where did the data come from? How would results shift under alternative assumptions?

AI tools can help organize sources and audit formulas, but citations and calculations remain your responsibility. Anything listed on your resume must be explainable by you; “the tool generated this output” cannot substitute for your judgment.

05 Investment volume alone cannot guide your job search

Greater bank lending to AI does not automatically translate into more new-grad headcount. How work is allocated across existing teams and outsourced to vendors shapes hiring, and news articles cannot answer those questions.

When applying, review each department’s mandate, experience requirements, graduation cohorts and location individually. For international students, work authorization and visa sponsorship need separate verification. Do not assume a role is suitable simply because the firm works on AI financing.

If you follow industry news but remain unsure whether to target banking, research or risk roles, you can review your resume and target roles with VINE Education. Together you can map your coursework, internships and projects to viable opportunities, identify supplementary industry research, and refine your resume and interview storytelling.

You do not need to chase every AI headline for recruiting prep. Pick one financing deal aligned with your target role, trace how the capital will be deployed and repaid, and articulate your view clearly. That research will serve you well.