[ 2025-12-21 22:34:36 ] | AUTHOR: Tanmay@Fourslash | CATEGORY: BUSINESS
TITLE: Newcomers Challenge Big Tech in AI Data Center Boom
// The AI data center market is seeing a surge of newcomers diversifying ownership from Big Tech, with massive projects in Italy and Canada highlighting economic risks and investment flows.
- • Global planned data center capacity reaches 145,000 MW, with new players accounting for 121,000 MW, surpassing Big Tech's 68,000 MW.
- • At least $178.5 billion in US data-center credit deals struck in 2025, part of over $6.57 trillion in global bond issuance.
- • Projects like Adriatic DC's 2,000 MW campus in Italy and Kevin O'Leary's Alberta park aim to create major AI hubs, financed largely by private capital.
AI Data Center Gold Rush Attracts Thousands of New Entrants
The dominance of Big Tech in AI infrastructure is eroding as thousands of newcomers with limited experience enter the fray, driving a global buildout projected to exceed $1 trillion and carrying broad economic risks.
Global live and planned data center capacity totals 145,000 megawatts (MW), according to analysis of industry data. New players account for 121,000 MW, outpacing Big Tech's 68,000 MW and financial firms' 44,000 MW. This diversification spreads risk but amplifies potential fallout if the AI boom falters, experts warn.
In Puglia, Italy, Adriatic DC, a one-year-old company led by CEO Lorenzo Avello, plans three data centers totaling at least 2,000 MW — enough to power about 1.5 million U.S. homes. The project, estimated at over 50 billion euros ($59 billion), aims to create a "Mediterranean AI hub" serving Europe, Asia and the Middle East. Avello, a former renewable energy developer, has secured land and electricity commitments, relying on private financing. He draws inspiration from the U.S. Stargate venture by OpenAI, Oracle and SoftBank, which pledges up to $500 billion in rural AI expansions.
Similar ambitions are unfolding in Canada. Shark Tank host Kevin O'Leary is developing what he calls "the world's largest AI data center industrial park" in Alberta, powered by natural gas and geothermal energy. O'Leary, who co-founded a software firm, predicts only a handful of such hyperscale projects will succeed amid hundreds of failures. He secured the site after scouting U.S. options, citing cheap power as a key advantage.
Financing Surge Fuels Global Expansion
Lenders have committed at least $178.5 billion to U.S. data-center projects this year, per compiled figures. Global bond issuance topped $6.57 trillion in 2025, propelled by tech giants like Oracle, Meta and Alphabet. JPMorgan Chase expects issuers to access all major debt markets to fund the infrastructure push.
Demand stems from AI leaders like OpenAI's Sam Altman and Nvidia's Jensen Huang, who foresee trillions in investments to reshape the economy. A Bloomberg Intelligence survey of over 600 C-suite executives ranked AI infrastructure access as a top concern across sectors.
Yet skeptics highlight bubble risks. Investor Michael Burry and others question the sustainability of deals where Nvidia and Microsoft invest in customers buying their products. Charles Fitzgerald, a data center analyst, doubts many projects will materialize, noting a lack of customers and the mismatch of private equity or celebrities in operating frontier-scale facilities.
Oaktree Capital's Howard Marks warns of overbuilding, with flexible leases allowing tech firms to exit contracts, leaving developers exposed.
Regional Shifts and Economic Implications
In Europe, new players could position southern Italy as the continent's largest AI hub. Adriatic DC's 2,000 MW dwarfs projects like Mistral AI's 1,400 MW in France and Ili Group's 600 MW in Spain. Puglia's sparse population, subsea cables and renewable potential make it ideal, Avello said.
The influx diversifies risk beyond tech but ties the global economy to AI's fate. A collapse could trigger casualties in equity, debt and unrelated sectors, analysts say. Avello remains optimistic, citing moves by governments and investors in the U.S., China and Saudi Arabia as evidence of AI's disruptive potential.
O'Leary echoes this confidence, predicting his project will rank among the top five hyperscalers. "I'm arrogant enough to say I will succeed," he said.
As supply lags demand today, the buildout proceeds. But long-term, excess capacity looms, potentially reshaping industries from energy to finance.
Tanmay is the founder of Fourslash, an AI-first research studio pioneering intelligent solutions for complex problems. A former tech journalist turned content marketing expert, he specializes in crypto, AI, blockchain, and emerging technologies.