Anthropic is preparing for an initial public offering that could value the AI company at more than $2 trillion, even as its prospectus shows rapidly rising infrastructure spending and widening losses, according to Reuters. The filing lays out Anthropic’s expectation that artificial intelligence could reshape the global economy while also illustrating the extraordinary capital required to compete at the frontier of AI.
Anthropic generated nearly $4.6 billion in revenue in 2025, representing approximately 12-fold year-over-year growth, according to the prospectus reviewed by Reuters.
That expansion came with substantial spending as the company invested heavily in computing infrastructure, cloud capacity, employees, and continued development of its Claude AI models.
Anthropic reported a net loss of nearly $42 billion for 2025, although the majority of that figure was not attributable to ordinary operating expenses.
About $34 billion of the net loss resulted from an accounting charge tied to an increase in the estimated value of financing instruments that could eventually convert into Anthropic shares.
On an operating basis, the company’s loss widened to approximately $8.06 billion in 2025 from $2.98 billion in 2024.
Anthropic’s spending on computing and infrastructure reached $7.33 billion in 2025, approximately three times the prior-year level.
Those expenses represented more than half of the company’s $12.65 billion in total operating expenses, underscoring how access to computing capacity has become one of the largest expenses for companies developing frontier AI systems.
The financial commitments extend well beyond the company’s latest fiscal year.
Anthropic disclosed plans involving approximately $518 billion of cloud, computing, and infrastructure obligations in coming years, illustrating the scale of resources it expects to require as it trains more advanced models and serves a growing customer base.
The IPO could establish one of the most significant public-market benchmarks yet for a pure-play artificial intelligence company.
A valuation above $2 trillion would be more than double Anthropic’s estimated valuation of approximately $965 billion in May 2026.
Anthropic was founded about five years ago by former OpenAI researchers and has become one of OpenAI’s largest competitors in generative AI.
Businesses, developers, and consumers use its Claude models for software development, research, writing, data analysis, and increasingly autonomous agent workflows.
The company’s prospectus also shows significant customer concentration.
According to Reuters, nearly one-quarter of Anthropic’s 2025 revenue came from two customers.
Anthropic warned that many of its largest customers are not locked into long-term contracts and could reduce or discontinue spending, creating potential revenue volatility even as overall adoption continues to grow.
Anthropic reported $20.28 billion in cash, cash equivalents, and short-term investments as of December 31, 2025.
The company has also received billions of dollars of strategic investment from technology companies including Amazon and Google, which simultaneously provide cloud infrastructure supporting the development and deployment of Claude models.
The prospectus also addresses risks associated with increasingly capable and autonomous AI systems.
Anthropic’s own research has shown that advanced models can behave unexpectedly in controlled evaluations, including scenarios involving code sabotage, fraud assistance, and information manipulation.
The disclosures highlight a tension facing frontier AI developers: companies are investing aggressively to release increasingly capable systems while simultaneously developing safeguards intended to reduce the possibility of harmful or unintended behavior.
CEO Dario Amodei has publicly advocated greater caution around the pace at which advanced AI capabilities are released, even as Anthropic competes intensely with OpenAI and other model developers.
Anthropic recently introduced Claude Opus 5.5, its latest flagship model, as competition accelerates across coding, agentic AI, reasoning, and enterprise applications.
The company’s competitors include OpenAI, Google’s AI operations, Meta, and xAI, while major technology companies continue spending heavily on the data centers and computing infrastructure required to support AI models.
The IPO would also provide public investors with more direct exposure to a frontier AI model developer rather than primarily gaining exposure through cloud providers, semiconductor companies, or diversified technology businesses.
Reuters previously reported that Anthropic’s public debut is likely to occur after the November 2026 U.S. midterm elections, citing people familiar with the matter. The timing remains subject to market conditions and the IPO process.
The potential offering follows a period in which investors have assigned extremely high private-market valuations to leading AI companies despite the massive amount of capital required to develop and operate their technologies.
Anthropic’s prospectus provides one of the clearest financial pictures yet of that tradeoff: revenue is expanding rapidly, but the company is simultaneously committing hundreds of billions of dollars to the infrastructure required to compete in AI at global scale.
