Stargate and the $500 Billion Question

On 21 January 2025, President Trump stood in the White House alongside Sam Altman, Masayoshi Son, and Larry Ellison to announce Project Stargate — a joint venture between OpenAI, SoftBank, and Oracle to invest US$500 billion in AI infrastructure across the United States over four years. The announcement was immediately met with a mixture of excitement, scepticism, and, in some quarters, disbelief. US$500 billion is roughly the annual GDP of Sweden. It is more than the entire global data centre industry invested in the previous five years combined. The question is not whether Stargate is significant — it clearly is — but what it actually means for the infrastructure investment landscape.

The first thing to note is that Stargate is a commitment, not a cheque. The US$500 billion figure represents a four-year investment target, of which US$100 billion is described as immediately available. The remainder depends on financing that has not yet been arranged, from investors who have not yet committed. This is not unusual for large infrastructure programmes — the US interstate highway system was also announced before the funding was fully secured — but it does mean that the headline number should be treated as an aspiration rather than a certainty.

The second thing to note is that Stargate's significance extends beyond its dollar value. By framing AI infrastructure as a matter of national industrial policy — explicitly analogous to the interstate highway system or the space programme — the announcement changes the political economy of AI infrastructure investment in the United States. It makes it harder for local governments to block data centre development on environmental or aesthetic grounds. It creates a presumption in favour of fast-tracking permits and grid connections. And it signals to the rest of the world that the United States intends to compete aggressively for AI infrastructure leadership.

For investors outside the United States, Stargate is both a challenge and an opportunity. The challenge is that it concentrates a significant portion of global AI infrastructure investment in a single country. The opportunity is that it validates the infrastructure thesis globally — and that the markets best positioned to serve AI demand in Asia Pacific, the Middle East, and Europe are not going to be served by data centres in Texas. At Lensbridge, we remain focused on the Asian infrastructure opportunity, which Stargate has, if anything, made more compelling.