Key highlights:
- Anthropic filed its IPO prospectus targeting a ~$2T valuation, revealing 2025 revenue of $4.6B alongside a ~$42B net loss
- Computing costs reached $7.33B in 2025 (the largest expense), with $518B in future cloud and infrastructure commitments
- The IPO is expected after November’s midterm elections, with co-founders pledging 80% of personal equity to charity while retaining 50.1% voting control via a Founder LLC
Anthropic is preparing for a potential $2 trillion valuation while committing to as much as $518 billion in future cloud computing and infrastructure spending, putting the enormous cost of scaling advanced AI under scrutiny ahead of a potential IPO.
The AI startup reported a net loss of nearly $42 billion in 2025 despite revenue rising twelvefold to about $4.6 billion, according to its IPO prospectus reviewed by Reuters.
The figures show the growing gap between Anthropic’s rapidly expanding AI business and the infrastructure costs required to support it. The company is committing billions of dollars to computing capacity as it competes with OpenAI and other AI companies for customers, data center resources, and access to advanced chips.
The spending commitments also raise the stakes for Anthropic’s expected public offering. A valuation approaching $2 trillion would require investors to place substantial value on the company’s future growth while accepting the enormous capital requirements attached to building increasingly powerful AI systems.
Anthropic’s prospectus provides a rare look at the economics behind that expansion, showing how quickly AI revenues can grow while infrastructure and model development costs continue to climb.
Anthropic is chasing a $2 trillion valuation; can its AI growth justify it?
Anthropic’s rapid revenue growth is the main argument behind its potential $2 trillion valuation, but the company’s financials also show how much it is spending to achieve that growth.
Revenue rose twelvefold to about $4.6 billion in 2025, according to its prospectus. Yet Anthropic recorded more than $8 billion in operating losses, excluding certain accounting adjustments, while total operating expenses reached about $12.65 billion.
Computing and related costs accounted for roughly $7.33 billion, making them the company’s largest expense. Anthropic has also committed to about $518 billion in future spending on cloud services, computing capacity, and other resources as it prepares to support further growth in demand for Claude.
Those commitments show why revenue growth alone may not be enough to support a valuation of $2 trillion. Anthropic needs to turn increasing demand for its models into revenue at a scale that can eventually offset the costs of serving that demand.
The company has secured access to computing resources through partnerships with Amazon, Google, and SpaceX, which have also invested billions of dollars in Anthropic. That backing gives the company more resources to continue expanding, but it also reflects the capital-intensive nature of its business.
Anthropic reported a net loss of about $42 billion in 2025, although the figure does not represent an equivalent amount of cash burned.
About $34 billion of the loss came from an increase in the estimated value of financing arrangements that could eventually convert into Anthropic shares. The accounting charge therefore had a major effect on reported earnings without representing the same amount of current cash spending.
Anthropic held about $20.28 billion in cash, cash equivalents, and short-term investments as of December 31, giving it substantial liquidity to fund its expansion.
The proposed valuation would nevertheless demand much more than continued revenue growth. At more than $2 trillion, Anthropic would be valued at more than twice its estimated $965 billion valuation in May and above SpaceX’s reported $1.77 trillion valuation following its recent offering.
That means investors would be placing substantial value on Anthropic’s future revenue and its ability to capture a larger share of the AI market.
There are already signs of concentration within that revenue base. Nearly one quarter of Anthropic’s 2025 revenue came from just two customers, while many major clients were not locked into long-term contracts and could reduce or end their spending.
The Anthropic IPO is highlighted by a rapidly growing business, substantial financial backing, and a large customer base, but also significant costs and commitments. The central valuation question is whether the company can maintain its exceptional revenue growth long enough for the economics of its AI business to catch up with the expectations embedded in a $2 trillion valuation.
Can Anthropic convince investors that its AI risks are under control?
Anthropic is also facing questions over the risks associated with the increasingly capable AI systems it is developing as the company prepares to enter the public markets.
The company’s prospectus warns that advanced AI systems could behave unpredictably, including manipulating information, assisting fraud, and engaging in other potentially harmful activities.
Chief Executive Dario Amodei has previously called for the industry to slow the development of advanced capabilities and strengthen safety measures.
Those concerns are becoming more significant as AI agents gain greater access to external systems, software, and digital tools.
Recent incidents involving AI agents have raised questions about how companies can prevent systems from taking unauthorized actions.
Anthropic’s own research has also examined the potential for models to display deceptive or harmful behavior under controlled conditions.
The risks could become an important consideration for investors as Anthropic plans to deploy increasingly capable systems at scale.
The IPO also comes as leading AI companies face growing scrutiny over the cost and risks of commercializing advanced models.
OpenAI has reportedly pushed its expected IPO toward 2027, with CEO Sam Altman previously describing a 2026 listing as potentially ill-advised amid increasing attention on AI safety.
Anthropic’s planned listing, expected after the November U.S. midterm elections, could therefore provide an early test of how public investors value AI companies that require substantial spending on computing infrastructure and model development.
The filing also outlines its proposed governance structure. Anthropic’s seven co-founders have pledged to donate 80% of their personal equity to charitable causes, while a proposed Founder LLC would retain control through a special share carrying 50.1% of voting power on key corporate matters.
As Anthropic moves toward an IPO, investors will need to assess its rapid revenue growth alongside infrastructure costs, customer concentration, governance arrangements, and the risks associated with increasingly capable AI systems.
Source:: Anthropic Plans $518B AI Infrastructure Spending as $42B Loss Raises IPO Stakes