OpenAI has a $852 billion valuation, $2 billion in monthly revenue, and a product enterprises are racing to adopt. It is also paying private-equity firms a guaranteed 17.5% annual return to join its enterprise deal — and one of the world's most sophisticated software investors already looked at the terms and walked away.
That contradiction is the story.
The company is racing to close a new joint venture, internally called DeployCo, by early May, according to the Financial Times. The Delaware-listed vehicle has TPG, Bain Capital, Advent International, Brookfield Asset Management, and Goanna Capital committing roughly $4 billion at a pre-money valuation of about $10 billion. OpenAI is putting up $500 million initially, with up to $1.5 billion total and an option for another $1 billion, and holds super-voting shares. Reuters reported April 22 that it could not immediately verify the FT account; OpenAI and the PE firms did not immediately respond to requests for comment.
The structure lets OpenAI absorb the high upfront costs of deploying engineers to customize its models for corporate clients, producing cleaner segment reporting ahead of a public offering. PE firms control hundreds of established companies and directly influence how those businesses budget for software. Lock in the channel now, the thinking goes, and the switching costs take care of the rest. "There's a big race to lock in as much enterprise, as many desks as possible," said Matt Kropp of Boston Consulting Group's AI unit.
Thoma Bravo, one of the world's largest software-focused buyout firms, decided not to participate. Managing partner Orlando Bravo questioned the long-term profit profile of the joint venture after internal discussions, a person familiar with the decision said. Thoma Bravo declined to comment.
OpenAI is projecting $14 billion in losses this year, even as its annualized revenue crossed $20 billion in 2025, a 233% year-over-year increase, according to Forbes. The company closed $122 billion in committed capital at an $852 billion post-money valuation in March, anchored by Amazon, Nvidia, and SoftBank. Anthropic raised $30 billion at a $380 billion valuation in February and has watched its enterprise market share grow from 18% to 29% in 2025, with eight of the Fortune 10 as customers. These are not companies at the same stage of the same race.
Anthropic is running a similar PE joint venture play, courting Blackstone, Hellman & Friedman, and Permira, but offering no guaranteed return. It does not need to. Its enterprise position is stronger, its unit economics are better, and its revenue trajectory is more credible to sophisticated software investors. OpenAI, burning $14 billion a year against $20 billion in revenue, is offering financial guarantees that its better-balanced rival refuses to make.
Other PE firms are expected to take smaller stakes without board seats or lead roles. Some large investors have noted that they already have direct commercial relationships with OpenAI and Anthropic without committing capital to a joint venture, making the incremental value of the formal structure unclear.
Both labs are positioning for IPO. OpenAI is targeting a public listing as early as this year, and Anthropic is running a parallel timeline. The enterprise JV is, in part, a race to demonstrate distribution depth before the roadshow begins. Whoever locks PE-backed enterprise first arrives at the IPO with contracted customer depth that a competing S-1 cannot easily replicate.
If OpenAI — the company that raised $122 billion to build transformative AI — is paying 17.5% guaranteed to private equity firms rather than reinvesting every dollar into the core technology, the question is not whether the enterprise deal will close. The question is what that guarantee says about where the actual returns in AI are being generated, and who is capturing them.
