Citi is eyeing the AI dealmaking frenzy as its moment to close the gap with Wall Street rivals.
In late February, the bank established an AI Infrastructure group, noting in an internal memo that the build-out could require an estimated $3 trillion in capital by 2030 — and that Citi wanted to build a “leadership position” in the space. The group has united leaders across the investment bank — from technology and communications to energy and power to real estate and crypto — to move faster on financing opportunities.
“This is our time,” said Achintya Mangla, who joined Citi in the fall of 2024 as head of financing in the investment bank after a 22-year run at JPMorgan. Mangla is a key architect of the new AI Infrastructure team — and he says Citi’s setting its aspirations high.
“No bank is an incumbent. No bank,” he told Business Insider in an interview. “We have the opportunity.”
For Citi, throwing their hat into that gauntlet is also a high-profile test of whether CEO Jane Fraser’s re-engineered version of the firm can compete for more “lead left” mandates — the top roles on financing and advisory deals that carry the most influence and fees — long dominated by its Wall Street rivals.
Other big banks are all in on data centers. Fred Turpin, the global chair of investment banking at JPMorgan — Mangla’s former workplace— recently told Business Insider that the build constitutes “the largest investment cycle in the history of capitalism.”
But Mangla is betting that Citi’s strategy will help it win market share from some of the most established players.
Shattering silos
Citi’s large-cap banking rivals have longer track records in data center financing and still dominate the market — but the margins are growing narrower.
Dealogic data from April shows that Citi ranked fifth for data center debt activity this year — up from sixth last year and eighth in 2024 — illustrating the bank’s progress. In overall US M&A activity over the past quarter, Citi ranked fourth behind Goldman, JPMorgan, and Morgan Stanley, respectively, according to Dealogic’s league table.
Since March 2025, Citi has handled more than $75 billion of data center construction financings, supporting roughly 6.1 gigawatts of IT capacity — roughly half of Con Edison’s forecast peak summer demand for 2025, a spokesperson for the bank said. One deal it worked on was Blue Owl and STACK Infrastructure’s $18 billion Stargate campus in New Mexico, which closed in 2025, for which it helped arrange financing.
Mangla, who previously ran equity capital markets at JPMorgan, argues that long-term success in the sector requires a new way of thinking about financing. On deals of this scale, banks must now evaluate a wide range of risks from power supply and land to construction execution to specialized GPU hardware to long-term contracts with customers like Meta or Microsoft.
With AI infrastructure projects moving so quickly, Mangla said handoffs between separate teams handling each leg of the financing process might create bottlenecks. Citi’s answer has been to bring those capabilities together into a single group. The aim is to move away from pushing a single product, like a bond or loan, and instead structure solutions across a range of asset classes.
“There is not one single person that can do all this,” he said. “What we really need is problem solving” and the ability to be “agnostic in providing a capital solution — whether it is debt, mezzanine, equity, or anything in the middle.”
The next ‘major player’?
The next step is generating results — and Mangla says the AI Infrastructure group has the right people in place to meet the challenge.
The main leadership group was complete, Mangla said, but the bank was open to adding select junior resources if exceptional talent materialized in the marketplace.
“We were humble and realistic about what we had to do,” Mangla said of the bank’s initial forays into the data center infrastructure financing space. “We didn’t go into client meetings saying we are number one.”
Brian Mulberry, the chief market strategist at Zacks Investment Management, said the setup of the AI Infrastructure team is emblematic of what he called “the last major step” of the bank’s yearslong turnaround effort.
“This would transcend them into a major player with the major money center banks in a way that they weren’t competing before,” Mulberry said. “It’s the last real big step for Jane to accomplish, to be able to say the turnaround is done.”
