Blackstone bets $30B on Japan AI data centres despite bubble warnings

Blackstone plans $30 billion investment in Japan AI data centres, Nikkei reports

June 23 (Reuters) - Blackstone (BX.N), opens new tab is planning to invest $30 billion in Japan's AI data centers over the next three to five years, its president and chief operating officer Jonathan Gray told Nikkei ⁠in a recent interview, the business daily reported on Tuesday. The world's largest alternative asset manager is in discussions to develop facilities exceeding 1 gigawatt in the country, the report said, citing Gray. Blackstone did not ⁠immediately respond to a Reuters request for comment. It also plans to accelerate its private equity investments ⁠in Japan, the company said. Earlier this month, Blackstone had raised $13.1 billion for its Asia ⁠private equity fund, exceeding its initial target and marking its ⁠largest such fundraise in the region. Reporting by Jasmeen Ara Shaikh in Bengaluru; Editing by Vijay Kishore Our Standards: The Thomson Reuters Trust Principles., opens new tab

Blackstone plans $30B for Japan AI data centres, and is unbothered by bubble talk

President Jonathan Gray told Nikkei the firm is in discussions to build facilities exceeding one gigawatt over the next three to five years. Blackstone plans to invest roughly $30 billion in artificial-intelligence data centres in Japan over the next three to five years, the firm's president and chief operating officer, Jonathan Gray, told Nikkei in an interview published on Monday. The world's largest alternative asset manager is in discussions to develop facilities with a combined capacity exceeding one gigawatt, a scale that would place its Japanese ambitions among the larger national buildouts anywhere. Gray framed the commitment as a bet on demand rather than a gamble on froth. Asked about the warnings that AI infrastructure spending has run ahead of the revenue meant to justify it, he argued that the risk of building too little computing capacity outweighs the risk of building too much. It is the position of a firm that has decided the cost of being early is smaller than the cost of being absent. The Japan plan does not arrive in isolation. Gray said Blackstone also intends to accelerate its private-equity investment in the country, where it has been an increasingly active buyer. Earlier this month the firm raised $13.1 billion for its latest Asia private-equity fund, exceeding its initial target and marking the largest such regional fundraise in its history. That capital and the data-centre push are, in Gray's telling, two expressions of the same conviction that Asia, and Japan in particular, is where the next leg of allocation belongs. Japan has become one of the more contested venues for AI infrastructure precisely because the constraints elsewhere are tightening. Power, land, and grid capacity have turned into the binding limits on data-centre construction, and Japan offers a stable grid, a government keen to attract investment, and proximity to the broader Asian compute market. SoftBank's domestic buildout, which extends from chips through to battery manufacturing, has already established Japan as a serious site for the kind of vertically planned infrastructure that hyperscale AI requires. A one-gigawatt threshold is a useful way to read the scale Gray is describing. It is roughly the output of a large power station, and it is the unit in which the most ambitious projects are now denominated. Blackstone has been moving in this direction for some time, having built a data-centre platform that ranks among the largest owned by any private investor, anchored by its 2021 acquisition of QTS. The Japanese facilities would extend that footprint into a market where few foreign managers have committed at comparable size. The bubble question is not one Gray can dismiss entirely, and he did not try to. The debate over whether AI valuations and capital expenditure have detached from fundamentals has grown louder as the numbers have grown larger, and the parallels to the dot-com build-out are difficult to ignore. Hyperscaler capital spending is approaching $660 billion to $690 billion this year, and a meaningful share of institutional investors now name an AI valuation crash as the single greatest risk to markets. Gray's answer is essentially structural: the compute will be needed, the question is only who owns it. That logic has competitors. China's drafted $295bn plan to knit its computing facilities into a single national grid runs on the same premise, that whoever controls the infrastructure controls the position, only pursued through sovereign debt rather than private capital. The American hyperscalers spend at a pace that makes even Blackstone's $30 billion look measured; Meta's Hyperion campus in Louisiana alone now carries a projected cost above $200 billion. Blackstone did not disclose the sites, partners, or financing for the Japanese facilities, and Gray characterised the projects as still in discussion rather than committed. The gigawatt-plus figure describes capacity under negotiation, not capacity under construction. What is firm is the direction. A manager that has spent the past few years assembling one of the largest data-centre portfolios in private hands has now named Japan as the next place it intends to build, and has said so without hedging on the demand it expects to meet.