Anthropic recorded a net loss of $80.7 billion in the first half of 2026, as revenue surged and spending on computing infrastructure climbed.

The San Francisco-based AI company reported a net loss of $13 billion during the same period a year earlier, according to a draft initial public offering prospectus reviewed by The Wall Street Journal. The vast majority of its latest loss, $72 billion, stemmed from accounting charges tied to changes in the value of convertible notes and other financial instruments issued to raise capital.

The disclosures come as Anthropic prepares for a potential blockbuster IPO that could value the company at as much as $2 trillion, while investors scrutinize the costs and financial risks of developing increasingly powerful AI models.

Anthropic generated about $16.3 billion in revenue in the six months ended June 30, up from $1.1 billion a year earlier, according to the draft prospectus. Meanwhile, spending on computing capacity and infrastructure climbed to nearly $9 billion, compared with $2.3 billion in the first half of 2025.

That six-month total exceeded the $7.3 billion Anthropic spent on computing and infrastructure throughout 2025, when it generated nearly $4.6 billion in revenue. For the full year, the company recorded a $42 billion net loss, with more than $34 billion stemming from fair-value adjustments to its liabilities, largely tied to previous fundraising.

Anthropic’s Computing Commitments Top $600 Billion

Anthropic’s rapid growth is driving massive long-term infrastructure commitments as it seeks enough computing power to train and operate its AI models.

Anthropic faces hundreds of billions of dollars in long-term infrastructure expenses, including cloud contracts that largely cannot be terminated early. The draft prospectus details planned payments of $111.1 billion to Alphabet’s Google, $110 billion to Amazon Web Services and $31.4 billion to Microsoft, even if the company does not use all the computing capacity covered by those agreements. 

Separately, Anthropic could pay as much as $86 billion to rent Nvidia-powered computing resources from providers Nscale and Lambda under agreements running through 2033 and 2034, respectively.

Anthropic’s long-term infrastructure spending could surpass $600 billion when its obligations under a Google chip-leasing arrangement involving Broadcom are included, according to the Wall Street Journal. The company is also discussing additional agreements that could push its total commitments up by another $150 billion.

The company warned that access to computing capacity remains uncertain, saying it has sought more capacity than some major cloud partners were willing to provide as they increasingly view Anthropic as a competitor.

AI Safety and Data Center Backlash Pose Risks

Anthropic also outlined potential legal and operational risks associated with its AI models, including the possibility that misaligned systems could cause serious harm.

The prospectus warned that models could exhibit self-preserving behavior to avoid being shut down or generate responses resembling blackmail to persuade people to let them continue operating. Anthropic said damages from a catastrophic AI-related incident could be so substantial that insurance might be unavailable or insufficient to cover the losses.

The disclosures come as independent AI evaluator METR questions whether Anthropic’s safety tests can reliably detect dangerous behavior. In a review, METR said the company’s risk assessment did not provide enough evidence to confidently rule out serious misalignment and called for more realistic testing of its monitoring systems.

Anthropic also identified "political resistance" to data center construction as a business risk in its prospectus. The company warned that opposition to new facilities has "become more organized" and that its public profile could "draw particular attention to projects associated with us."

The prospectus also detailed executive compensation. CEO Dario Amodei received about $18 million in total compensation in 2025, while President Daniela Amodei received approximately $16.5 million, largely through stock and option awards.

Anthropic’s co-founders have pledged to dedicate 80% of their personal company equity to philanthropy. The company also plans to commit 5% of its shares to a new public benefit fund, according to the report.

Photo: RixAiArt / Shutterstock – ek