A report describes prospectus figures showing steep losses alongside enormous planned outlays on AI infrastructure.
Anthropic’s initial public offering prospectus reportedly discloses a $42 billion loss and spending plans reaching $518 billion, according to a report from CryptoBriefing published on September 28. The figures, as described, would mark one of the largest disclosed loss-and-spending profiles ever attached to a company preparing for a public listing.
Anthropic, known for its Claude family of AI models, has positioned itself as a leading competitor to OpenAI and other large language model developers. Companies in this category have consistently reported heavy losses tied to the cost of training and running large models. Compute, data center capacity, and specialized chips represent the bulk of these expenses.
The reported $518 billion spending figure would reflect the scale of infrastructure investment AI labs say is required to remain competitive. Training frontier models demands vast clusters of graphics processing units, long-term power contracts, and custom data center buildouts. These commitments often span years and are booked well ahead of revenue realization.
A $42 billion loss figure, if confirmed through the prospectus itself, would place Anthropic’s disclosed losses well above typical pre-IPO benchmarks for technology companies. Loss-making IPOs are not unusual in fast-growing tech sectors, but losses of this magnitude are rare outside capital-intensive industries like semiconductors or telecommunications infrastructure.
The prospectus disclosure, as reported, arrives amid broader scrutiny of AI industry spending commitments. Investors and analysts have increasingly questioned whether current revenue growth at major AI labs can eventually justify the scale of capital being deployed. Some have compared the spending trajectory to earlier infrastructure buildouts in telecommunications and cloud computing, where early losses preceded later profitability.
Because the report cited here is the only account currently describing these specific figures, readers should treat the numbers as preliminary pending confirmation through official regulatory filings. Prospectus documents filed with securities regulators typically undergo review and amendment before a company’s shares begin trading, and disclosed figures can shift during that process.
Sources disagree on this story
This article was published before the reports below were compared. The reporting above stands; what follows is where the published accounts do not agree.
CryptoBriefing says Anthropic could complete its IPO before the US midterm elections, while TechCentral and CoinDesk report the listing is expected to be pushed to after the midterms.
What all sources agree on
- Anthropic recorded a net loss of nearly $42 billion in 2025.
- Anthropic plans to spend $518 billion on cloud, computing and infrastructure.
- The IPO could value Anthropic at more than $2 trillion.
- Anthropic's May 2026 funding round valued the company at $965 billion.
Where the reports disagree
1Whether the IPO could complete before or after the US midterm elections
Anthropic is targeting a public listing on Nasdaq as early as mid-October 2026, potentially completing the process before the US midterm elections.
Anthropic’s public market debut is likely to be pushed to after the US midterm elections in November, Reuters previously reported.
Its public listing is likely to come after the November U.S. midterm elections and could value the company at more than $2 trillion, Reuters reported.
What would settle it: Anthropic's SEC S-1 filing or a public statement from the company confirming its intended listing date.
What to make of it
Treat the $42 billion loss, $518 billion spending plan, and $2 trillion valuation target as established across all three reports. The IPO timing relative to the midterm elections is unresolved and should not be treated as settled until Anthropic files its public S-1 or announces a listing date.
Market Impact
If the reported figures hold, Anthropic's IPO could become a reference point for how markets price loss-making AI companies with enormous capital needs. Public investors weighing exposure to frontier AI development would be asked to accept losses and spending commitments far larger than those typical of prior technology listings.
Broader market sentiment around AI infrastructure spending also affects adjacent sectors, including chipmakers, data center operators, and crypto-linked projects tied to decentralized compute or GPU access. A high-profile IPO disclosing losses and spending at this scale could sharpen debate over whether AI capital expenditure is outpacing near-term revenue generation across the industry.
The reported figures underscore how costly frontier AI development has become, even for well-capitalized firms preparing to go public. Confirmation through official filings will determine whether these numbers hold as Anthropic's IPO process advances.
Frequently Asked Questions
What did the report say about Anthropic's IPO prospectus?
CryptoBriefing reported that Anthropic's IPO prospectus discloses a $42 billion loss and spending plans totaling $518 billion.
Has Anthropic confirmed these figures directly?
The figures come from a report describing the prospectus, and official confirmation would require review of the filed regulatory documents themselves.
Why would an AI company report such large spending figures?
Training and running large AI models requires extensive computing infrastructure, including data centers, chips, and long-term power agreements, which drive up capital costs.
How does this compare to other tech company IPOs?
Loss-making IPOs occur regularly in fast-growing tech sectors, but losses and spending commitments of this reported scale would be unusually large even by industry standards.
