Jas Khara, who leads Blackstone Group LP's dedicated AI investment team, said on Wednesday that building out AI infrastructure will require enormous amounts of capital, but the firm is still assessing which types of investors will ultimately step in to provide that massive financing.

He added that the large investment institution is also watching whether debt issued by companies such as OpenAI and Anthropic offers attractive investment potential. Blackstone Group LP is itself deeply involved in AI infrastructure development: it is working with Google on a neocloud project, leasing Google's tensor processing unit (TPU) computing power, while the group's credit and insurance divisions are also making related investments directly.

Khara said that although Blackstone Group LP has clearly mapped out market demand for AI computing power, the types of debt investors suited to meeting that demand are still evolving. At The Information's "AI Agenda" event in San Francisco, an audience member asked whether financing costs would have to rise to support AI infrastructure. He responded: "We haven't figured out who will take on all of the debt. We have some guesses, but honestly, we don't know where all the money will come from."

Khara outlined a macro financing framework, listing three types of financing channels that could play a role: the public bond market for investment-grade companies, the private credit market, and corporate bonds issued by OpenAI and Anthropic after they go public. "A lot depends on whether they can successfully list, and whether they can subsequently achieve investment-grade credit ratings," Khara said. He noted that if the price is right, Blackstone Group LP might also acquire debt issued by AI labs. "But we are still watching how things develop, looking at both the companies' business progress and the level of debt pricing."