OpenAI, Anthropic face $852B valuation test as IPO spending builds

OpenAI and Anthropic face mounting pressure to keep funding frontier-model development as both approach public listings, even as circular financing structures and unproven profitability complicate the valuations investors will be asked to underwrite.

"Making accurate valuations of these two AI labs is difficult because of the circular financing of the AI trade, and because of challenges in assessing profitability and the need to maintain frontier AI status," Rupert Mitchell, author of Blind Squirrel Macro, said.

OpenAI completed a $7 billion employee stock buyback at an $852 billion valuation, matching its March financing round that raised $122 billion. The company generated roughly $5.7 billion of revenue in the first quarter but burned about $3.7 billion during the period, according to internal documents reported by The Information. OpenAI has already missed some internal revenue and user-growth targets this year as competition from Anthropic intensifies.

The stakes extend beyond the two labs. Anthropic's valuation has surpassed OpenAI's and it is expected to complete its IPO first, while OpenAI secretly filed for its own listing in June. CEO Sam Altman has told employees that technological developments could make delaying an IPO advantageous, giving management room to wait for stronger financial conditions before defending an $852 billion valuation in public markets.

Circular Financing Clouds the Valuation Picture

The AI trade's circular financing — where hyperscalers and investors fund the very labs whose compute demand justifies their own AI infrastructure spending — makes it hard to pin down what these companies are actually worth. Microsoft, a major OpenAI backer, and Google, which has underwritten Anthropic, both have a direct interest in keeping the labs funded. A slowdown at either lab could ripple through the broader AI capital cycle, since the same investors underwriting the labs are also funding the data centers and chips that power them.

This interdependence cuts both ways. OpenAI's $122 billion March round drew from major tech companies and venture capital firms, giving it a strong balance sheet for the $7 billion buyback it funded entirely from its own cash. Unlike earlier employee liquidity programs that relied on outside investors such as Thrive Capital and SoftBank Group, the latest tender offer required no external capital — a sign of financial strength, but also a reminder that the company is spending heavily to retain talent ahead of a listing.

Anthropic's rise has reshaped the competitive order. Once regarded as a follower in the AI race, its software has gained broad market recognition, its valuation now exceeds OpenAI's, and it is expected to reach public markets first. That puts OpenAI under twin pressure: maintaining a technological edge while accelerating its capital-markets strategy so it does not fall behind in the IPO window.

Profitability Remains the Open Question

OpenAI's cash burn of $3.7 billion against $5.7 billion in first-quarter revenue shows the gap between growth and profitability. Compute costs, which consume the bulk of spending, will need to fall as a percentage of sales for the company to approach public markets from a stronger position. Persistent losses or further competitive pressure would make defending an $852 billion valuation considerably harder.

The compute bill is the crux. Training and operating frontier models requires enormous computing infrastructure, and the cost of that infrastructure is what keeps both labs in a spending race. As long as the two companies compete to release the most capable models, their capital expenditures will stay elevated, making near-term profitability a distant prospect.

For investors, the question is whether the AI trade's valuations can hold once these companies face quarterly scrutiny. If revenue accelerates while compute costs fall, OpenAI and Anthropic could justify their price tags. If not, the circular financing that has propped up the sector could unwind, pressuring AI-related equities across the board. The two labs' listings will test whether the market's enthusiasm for frontier AI can survive the discipline of public-market reporting.

This article is for informational purposes only and does not constitute investment advice.