Anthropic’s confidential IPO prospectus has revealed the scale of the company’s rapid expansion, including nearly $4.6 billion in revenue during 2025 and a reported net loss of $42 billion. The financial details, first reported by Reuters on September 28, 2026, offer a closer look at the costs behind one of the leading developers of advanced artificial intelligence models.
The company’s revenue grew approximately 12-fold in 2025, reflecting rising demand for its Claude AI products and services. However, the growth came alongside a widening operating loss of about $8.06 billion, compared with $2.98 billion in 2024, showing that higher sales have not yet translated into operating profitability.
Anthropic’s prospectus also outlines $518 billion in planned cloud, computing and infrastructure obligations over the coming years. The commitments reflect the resources required to train, develop and serve increasingly capable AI models, as competition intensifies among major technology companies to secure computing capacity.
Anthropic’s Revenue Surges as AI Infrastructure Costs Rise
The reported $42 billion net loss requires further context. According to Reuters, approximately $34 billion of the figure came from an accounting charge related to the estimated value of financing instruments that could eventually convert into Anthropic shares. This charge was not equivalent to cash spent on operations, making the operating loss a more direct measure of the company’s underlying business performance.
Anthropic spent $7.33 billion on computing and infrastructure in 2025, roughly three times its spending in 2024. The amount represented more than half of the company’s total operating expenses of $12.65 billion, underlining the importance of computing resources to its business model.
The company’s infrastructure commitments could support further expansion as demand for AI tools grows. Training advanced models requires large computing clusters, while serving customers also creates ongoing costs associated with processing requests, maintaining systems and expanding capacity. Anthropic’s planned spending therefore reflects both model development and the infrastructure needed to operate its products at scale.
Despite its losses, Anthropic held $20.28 billion in cash, cash equivalents and short-term investments as of December 31, 2025, according to the prospectus. This provides a measure of its available financial resources, although the company’s future funding needs will depend on its spending, revenue growth and ability to secure additional capital.
The prospectus also highlights customer concentration as a business risk. Nearly a quarter of Anthropic’s 2025 revenue reportedly came from just two customers. The company warned that many of its largest clients do not have long-term contracts and could reduce or stop spending, leaving revenue exposed to changes in customer demand.
Anthropic’s Potential $2 Trillion IPO Faces Investor Scrutiny
Anthropic is reportedly considering an IPO valuation of more than $2 trillion, which would be more than double its estimated $965 billion valuation in May 2026. Such a listing would place the company among the most highly valued technology businesses and give public-market investors direct exposure to a leading AI developer.
The proposed valuation comes as Anthropic competes with OpenAI, Google, Meta and other companies for customers, talent and access to computing infrastructure. Its commercial growth has attracted substantial interest from investors and major technology partners, including Amazon and Google, which have provided both capital and cloud services.
However, the prospectus presents a more complex picture than revenue growth alone. The combination of rising operating losses, significant infrastructure commitments and customer concentration creates financial risks that prospective investors will need to assess. The scale of future spending also means that access to capital and the ability to turn AI demand into sustainable revenue could remain important factors in the company’s long-term outlook.
Related: Anthropic Says Claude Discovers CRISPR-Like Enzyme System Hidden in Bacteriophage DNA
The filing also discusses risks associated with increasingly capable AI systems. Reuters reported that Anthropic’s research has identified potentially harmful behavior in controlled tests, including models that could sabotage code, assist fraud or manipulate information. These concerns add a further dimension to the company’s public-market plans, as investors consider both the commercial potential and the risks associated with advanced AI.
Anthropic CEO Dario Amodei has previously called for a more cautious approach to releasing increasingly powerful AI capabilities. At the same time, the company continues to compete in a fast-moving market where developers are investing heavily in new models and infrastructure. Balancing safety measures with commercial expansion remains a significant challenge for the industry.
The timing and final terms of Anthropic’s IPO remain subject to market conditions and regulatory processes. If the company proceeds with a public listing, its financial disclosures could become an important reference point for assessing the economics of frontier AI development. For now, the prospectus illustrates a business experiencing exceptional revenue growth while taking on substantial costs and financial commitments to compete in the global AI market.
About The Author
