TEMPO.CO, Jakarta - Wall Street is emerging as one of the biggest financial winners of the global artificial intelligence (AI) boom, pouring billions of dollars into data centers and creating new financing models to fuel the industry's rapid expansion. At the same time, however, analysts are warning that soaring valuations, speculative investments, and intensifying competition—particularly from China—could expose investors to significant market risks.
According to The Wall Street Journal and cited by Seoul Economic Daily, the world's largest asset manager, BlackRock, is leading a debt financing package worth at least US$12 billion for a massive AI data center project in El Paso, Texas, backed by Meta Platforms. BlackRock, through its infrastructure and private credit divisions, owns an 80 percent stake in the project, while Meta holds the remaining 20 percent and is expected to use the facility, which will have a planned capacity of about one gigawatt.
JPMorgan Chase and Morgan Stanley are leading the financing effort while approaching additional investors to participate in the project.
The investment reflects a broader surge in AI infrastructure spending. Over the past year, BlackRock has committed tens of billions of dollars to AI-related data centers, including its acquisition of Aligned Data Centers in a US$40 billion deal. It also participated in a record US$27 billion private debt issuance that helped finance Meta's Louisiana data center, purchasing more than US$3 billion in bonds.
According to the Financial Times, Morgan Stanley has become one of the leading architects behind the financial structures supporting AI expansion, advising on multibillion-dollar financing packages for companies including Meta, Broadcom, TeraWulf, and cloud computing provider CoreWeave.
The investment bank recently arranged a US$3.1 billion syndicated loan to finance CoreWeave's purchase of Nvidia graphics processing units (GPUs), creating one of the first financing structures that separates data center infrastructure from the AI chips powering it.
Morgan Stanley executive Graham compared the arrangement to buying a luxury car.
"The chips are the Ferrari. You need a place to park them, so you need data centers to house the chips."
The financing model has attracted enormous investor interest. A US$3.1 billion loan reportedly generated nearly US$20 billion in demand, highlighting investors' appetite for AI infrastructure backed by long-term computing contracts from major technology companies.
Those technology giants—often referred to as "hyperscalers"—including Google, Amazon, Meta, and Microsoft, have become central to the AI financing ecosystem. Their long-term lease agreements and computing contracts significantly reduce financing costs for developers building data centers.
Morgan Stanley estimates that roughly US$10 trillion could be invested globally in AI infrastructure over the coming years.
However, while capital continues flowing into AI projects, some analysts warn that investor enthusiasm is beginning to resemble previous technology bubbles.
As reported by ABC News, concerns intensified after Chinese AI startup Moonshot introduced its Kimi K3 model, which reportedly rivals leading systems developed by OpenAI and Anthropic despite being built at a fraction of the cost.
The development raises the prospect of a price war in AI services that could undermine the enormous spending commitments made by major U.S. technology companies.
The concerns have been amplified by the disappointing market debut of SpaceX, Elon Musk's latest publicly listed company, whose shares fell below their initial offering price just one month after listing.
The decline has prompted renewed scrutiny of AI-related valuations, particularly after Wall Street analysts overwhelmingly recommended buying the stock despite the company's continuing losses.
According to the report, 18 of the 19 analysts covering SpaceX rate the stock as a "buy," with none recommending investors sell despite its recent decline.
Some investment banks have attached ambitious price targets to the company. Goldman Sachs projected shares could reach US$205 within a year, while Morgan Stanley estimated US$300. Those valuations imply prices exceeding more than 100 times the company's annual revenue rather than its profits.
Unlike many established technology companies, SpaceX remains unprofitable while pursuing costly ambitions across multiple businesses, including satellite internet, social media, rocket launches, and artificial intelligence.
The combination of soaring valuations, record-breaking investment flows, and increasingly aggressive financing structures has prompted comparisons with previous speculative periods such as the dot-com boom and the railroad investment bubble.
While AI continues to transform technology and attract unprecedented levels of capital, analysts say the industry's future may ultimately depend on whether massive infrastructure spending can generate sustainable returns before competition and lower-cost alternatives begin reshaping the market.
Read: Why Christopher Nolan Calls AI a 'Trojan Horse'
Click here to get the latest news updates from Tempo on Google News
