Anthropic has formally opened its books in an S-1 filing, revealing a financial picture defined by exponential revenue growth, widening operating losses, and a stark warning about the potential dangers of artificial intelligence.
What Happened
According to a review of the prospectus by the Financial Times and Reuters, Anthropic warned potential investors that its own technology could pose "existential risks to humanity." The company reportedly distributed the S-1 filing to a small group of partners recently. Nearly a third of the lengthy document is dedicated to risk factors, including the possibility that increasingly advanced AI models could manipulate, blackmail, or behave in unpredictable ways.
The filing highlights the enormous cost of the race to build capable models. Revenue grew twelvefold in 2025 to nearly $4.6 billion, as reported by Reuters. However, the operating loss widened from $2.98 billion to $8.06 billion. Anthropic spent $7.33 billion on compute and infrastructure alone in 2025, a figure that is three times the previous year's spending and accounts for more than half of total operating costs.
The company also disclosed a net loss of roughly $42 billion, approximately $34 billion of which stems from an accounting charge reflecting the higher estimated value of financing that could later convert into stock, rather than cash spent on operations. Additionally, Anthropic noted a heavy dependence on a small number of buyers, with just two customers accounting for nearly a quarter of revenue in 2025. Many large customers are not locked into long-term contracts.
Looking forward, Anthropic plans to spend a total of $518 billion on cloud, compute, and infrastructure commitments over the coming years. The prospectus states the company expects AI to reshape the global economy more deeply than industrialization, electricity, and the internet did. In the second quarter of 2026, Anthropic generated $11.5 billion in revenue, according to the FT, and is on track for its second consecutive quarter of operating profit on an adjusted basis.
Why It Matters
Anthropic's path to the public markets is being watched closely as a potential benchmark for the entire AI industry. Backers believe a valuation above $2 trillion is possible, according to the Financial Times. This would be more than double the $965 billion valuation from May and would exceed the roughly $1.8 trillion valuation of SpaceX when it went public in June.
Analysts expect the first AI company to list on the stock market to set valuation benchmarks for the sector, including rival OpenAI, which confidentially filed for an IPO in June. Reuters sources indicate that Anthropic's debut likely won't occur until November, after the US midterm elections. The filing's emphasis on "existential risks" and unpredictable model behavior also signals a shift in how frontier AI developers communicate safety concerns to institutional investors, moving beyond technical papers to formal legal disclosures.
The Bottom Line
Anthropic's S-1 filing reveals a company with explosive revenue growth and massive infrastructure spending, balanced against significant operating losses and concentrated customer risk. By explicitly citing existential risks and unpredictable AI behavior in its risk factors, Anthropic is framing its public offering not just as a financial event, but as a pivotal moment for industry valuation and safety disclosure standards.