Anthropic’s New Prospectus Shows Revenue Is Growing Fast, So Are Its Compute Costs, Losses And AI Risks: Report

• Anthropic generated $16.3 billion in revenue during the first half of 2026—up dramatically from $1.1 billion a year prior.
• The Claude developer also posted an $80.7 billion net loss during the period, primarily driven by non-cash accounting adjustments.
• The artificial intelligence giant faces over $600 billion in future cloud and computing commitments, with potential negotiations threatening to add another $150 billion.

Claude maker Anthropic is gearing up for a blockbuster initial public offering that could value the firm at up to $2 trillion, according to confidential draft prospectus documents viewed by The Wall Street Journal. 

The filing showcases Anthropic’s revenues expanding at a heightened pace, but that is being accompanied by soaring compute costs, financial commitments, and valuation adjustments. 

Read Next

SpaceX, Tesla Form Independent Committees To Explore Merger: Elon Musk Says Deal ‘Not A Certainty’

Anan Ashraf

During the first six months of 2026, the San Francisco-based firm earned $16.3 billion in revenue, a massive leap from $1.1 billion in the same period in 2025, WSJ reported, citing the new draft prospectus it exclusively viewed. 

However, the company also reported a staggering net loss of $80.7 billion for the first half of the year. About $72 billion of this figure stemmed from non-cash accounting charges tied to valuation adjustments on convertible notes and other financing instruments issued to secure early capital. By comparison, Anthropic lost $42 billion in all of 2025.

Hundreds Of Billions In Commitments

The firm spent nearly $9 billion on infrastructure and computing capacity in the first half of 2026 alone, surpassing its entire $7.3 billion infrastructure expenditure for all of 2025.

To secure its future training and operational capacity, Anthropic has locked itself into roughly $450 billion in noncancelable cloud agreements spanning the next seven to ten years. 

Commitments include $111 billion allocated for Google cloud computing contracts, $110 billion committed to Amazon Web Services, $31 billion earmarked for Microsoft services, $86 billion for renting Nvidia graphics processing units via smaller cloud providers like Nscale and Lambda, and over $160 billion dedicated to leasing Google chips through Broadcom partnerships.

The company also noted that major cloud partners including Amazon, Google, and Microsoft are increasingly viewing Anthropic as a competitor and have been unwilling to provide the totality of the capacity it requires, leading to individual compute spends and third-party commitments that are pushing costs higher.

Unique AI Risks And Operational Vulnerabilities

Nearly a third of Anthropic's draft prospectus was dedicated to detailing risk factors, highlighting the distinct challenges facing advanced frontier AI developers. One of the top mentioned concerns was AI models exhibiting self-preserving behaviors or generating manipulative responses, including blackmail to avoid being shut down by human operators.

Anthropic also warned that it could face immense legal and financial liabilities for damages caused by its models, ranging from compromised data security incidents to catastrophic real-world disruptions. 

Retail sentiment on Stocktwits was ‘extremely bullish’ for Anthropic, with ‘high’ message volumes. 

For updates and corrections, email newsroom[at]stocktwits[dot]com.