Anthropic faces $42bn loss and $518bn infrastructure bill, IPO filing reveals
AI company prepares for potential $2tn stock market debut
Anthropic recorded a net loss of $42 billion in 2025, despite its revenue increasing 12-fold to nearly $4.6 billion, according to a draft prospectus for its planned initial public offering (IPO) reviewed by Reuters.
Anthropic, the developer of the Claude chatbot, submitted a confidential IPO filing to the US Securities and Exchange Commission (SEC) in June, although the prospectus has yet to be made public.
The company is reportedly targeting a valuation of more than $2 trillion, more than double the $965 billion valuation it achieved in May.
Its listing is expected to take place after November's US midterm elections, according to Reuters.
Accounting charges and soaring infrastructure costs
About $34 billion of Anthropic's net loss stemmed from an accounting charge reflecting an increase in the estimated value of financing that could eventually be converted into company shares.
The charge did not represent money spent on operating the business.
However, Anthropic's operating loss, which excludes the accounting adjustment, increased to $8.06 billion in 2025, up from $2.98 billion a year earlier.
Its operating expenses totalled $12.65 billion, including $7.33 billion spent on computing and infrastructure - three times the 2024 figure.
Despite its rapid growth, Anthropic warned that many of its biggest customers do not have long-term contracts and could reduce or discontinue their spending.
The prospectus also reveals substantial financial commitments to secure the computing power needed to develop and operate its AI models.
Anthropic expects to spend at least $518 billion over the next decade on infrastructure supplied by six partners. About 80% of that amount is non-cancellable or must be paid regardless of how much capacity the company uses.
Its largest commitments include $111.1 billion to Google, $110 billion to Amazon and $31.4 billion to Microsoft, under agreements lasting between seven and 10 years.
Anthropic also has around $161.2 billion in equipment leases associated with Broadcom, which has sought to raise $60 billion in debt to fund its chip operations.
AI safety warnings and concentration of control
The prospectus reportedly devotes about 80 of its 261 pages to risk factors, including the potential consequences of increasingly powerful AI systems.
Anthropic chief executive Dario Amodei has been among the prominent voices calling for a slower pace of development of advanced AI, a position he has described as "pacing the frontier".
Addressing the UN Security Council last week, Amodei warned that AI could threaten humanity, describing it as "the most important global security issue facing the world today".
Concerns have intensified following several high-profile cybersecurity incidents involving AI agents breaching external systems.
Jack Collier, chief growth officer at io.net, said Anthropic's disclosures raised questions about the concentration of power within the AI industry.
"The debate over who controls AI infrastructure has even reached the Vatican, with Pope Leo XIV publicly challenging Nvidia CEO Jensen Huang's claim that the industry can regulate itself without government oversight."
"The most pressing AI risk this decade isn't dangerous models - it's that only a handful of companies control those models and the technology and infrastructure behind them. A fairer industry would provide broader, more equitable access to GPUs, rather than increasingly long leases from the same few landlords."
IPO race could set benchmark for AI industry
Anthropic's financial disclosures come as it prepares to compete with OpenAI for investor attention in what could become a defining period for the AI industry.
OpenAI also confidentially filed for an IPO in June and is reportedly expected to go public by early 2027.
The two firms are vying for corporate clients, skilled workers and political influence in Washington, while competing with other major players, including Google's parent Alphabet, Meta and Elon Musk's xAI.
Analysts believe the first major AI firm to list publicly could set valuation benchmarks for the industry and draw in investors seeking a direct way to gain exposure to the technology.