Morning Edition · Wednesday, September 30, 2026Published at 2:34 AM EDT · New York
The filing reviewed by Reuters shows a $42 billion net loss for 2025, about $34 billion of it a non-cash charge tied to convertible financing, and $518 billion in planned compute and infrastructure spending.

A draft of Anthropic's prospectus for an initial public offering reached Reuters on Monday, and it contains the most detailed public accounting yet of what it costs to build a frontier AI lab. The company reported $4.6 billion of 2025 revenue against roughly $13 billion of operating expenses, an operating loss above $8 billion, and a net loss near $42 billion.
The gap between the operating loss and the net loss is significant. About $34 billion of the net loss is a non-cash charge tied to convertible financing that was revalued upward as Anthropic's own valuation rose. No cash left the company because of that item. Anthropic reported $20.28 billion in cash, against a stated plan to spend $518 billion on cloud computing, compute and infrastructure in the coming years.
The risk section is the other unusual feature of the filing. Anthropic devoted 80 pages to risk factors against 48 pages describing the business, and stated that the technology it is building could pose a catastrophic or existential risk to humanity. That language matches the lab's public position, including the economic scenario model its economics team published this month, and it is now a disclosure item that public-market investors must weigh.
A listing could come as soon as November at a valuation above $2 trillion, according to the reporting. The document is a draft, and neither the valuation nor the timing is final.
Anthropic and its existing private investors, since a leaked draft establishes a $2 trillion reference point before any banker sets a range, while cloud partners holding the forward compute commitments gain validation of the spending plan.
Part of a tracked trend
AI Hype Cycles and Funding Narratives
As capital floods AI, the narratives labs use to raise money and shape rules face growing public scrutiny, and the market increasingly separates verifiable capability and revenue from rhetoric on both the bullish and the cautionary side.
Start a discussion in Townsquare.
More from this edition
The financial figures match independent readings of the leaked document, yet the article omits that about a quarter of revenue came from two customers on contracts that are not long-term, and a leak of a draft prospectus is itself a form of pre-marketing whose timing serves the issuer.
An open-source-intelligence read of how likely this story is true with its real nuance, not a judgment of any outlet. It assesses the claim, weighing independent and adversarial reporting. How we label confidence.
What this means
The mechanism is disclosure. A private lab's cost structure becomes an audited public benchmark that every other AI capital raise gets measured against. Exposed are Anthropic's private-round investors, whose valuations now face a public comparison, the cloud providers holding those $518 billion in forward commitments, and OpenAI, which will face the same questions about revenue relative to compute spending whenever it approaches public markets. The key is separating the accounting from the operating performance: an $8 billion operating loss on $4.6 billion of revenue describes a materially different company than a $42 billion headline loss does. The likelier outcome over the coming months is that gross margin and compute-commitment figures, not the headline valuation, determine how the offering is received, and that rival labs speed up their own disclosures to shape the comparison. This would be proved wrong if the listing is pulled or priced sharply lower than expected, which would indicate that public investors are rejecting the compute-spending strategy rather than the accounting method.
What to watch
Observations to monitor, not financial advice.
Synthesized from: Polylog editors · Anthropic Research (Economics)
Comments
0No comments yet.