San Francisco: Anthropic, the company behind Claude AI, has disclosed a $42 billion net loss for 2025 while outlining plans involving as much as $518 billion in future cloud, computing and infrastructure commitments, according to its IPO prospectus reviewed by Reuters. The filing highlights both the rapid expansion of the artificial intelligence company and the enormous financial costs associated with developing and operating increasingly capable AI systems.
Anthropic is also seeking a public-market valuation of more than $2 trillion through its planned initial public offering, according to the prospectus and reporting around the proposed listing. That would be more than twice the roughly $965 billion valuation the company reached in its May 2026 funding round.
The proposed listing comes as the AI industry enters a new phase in which companies are seeking to turn rapid revenue growth into sustainable businesses while spending heavily on computing capacity.
Revenue surges while losses remain substantial
Anthropic’s revenue increased about 12-fold in 2025 to nearly $4.6 billion, according to the IPO documents. Despite that growth, the company recorded an operating loss of about $8.06 billion, compared with $2.98 billion in 2024, excluding certain liability-related writedowns.
The headline net loss of around $42 billion requires additional context. Approximately $34 billion of the loss was an accounting charge associated with the estimated value of financing instruments that could eventually convert into Anthropic shares. It therefore does not represent $34 billion of cash spent on operating the company.
Anthropic nevertheless spent heavily on the infrastructure required to develop and operate its AI models.
The company spent about $7.33 billion on computing and infrastructure during 2025, more than three times its spending in the previous year. Anthropic ended December 2025 with approximately $20.28 billion in cash and short-term investments.
$518 billion infrastructure commitment
One of the most striking figures in the prospectus is Anthropic’s planned commitment of up to $518 billion towards cloud computing, computing capacity and infrastructure in the coming years.
The commitment reflects the capital-intensive nature of frontier AI development. Training and operating increasingly sophisticated models requires substantial computing resources, while demand for AI services is also pushing companies to secure long-term access to data centres and cloud capacity.
Anthropic’s financial disclosures therefore illustrate the gap between its rapid revenue growth and the infrastructure spending required to support that growth.
The company is betting that continued expansion of AI adoption will generate enough demand to justify those enormous commitments.
IPO could value Anthropic above $2 trillion
Anthropic’s planned stock-market listing could value the company at more than $2 trillion, according to the prospectus and reports surrounding the offering.
Such a valuation would represent a substantial increase from the company’s approximately $965 billion valuation following its May funding round.
The proposed IPO would also become an important public-market test for the valuation of major AI companies. A successful listing at that scale could provide investors with a direct market benchmark for one of the leading private AI labs.
However, the timing of the offering remains subject to change. Reuters has reported that the listing is likely to take place after the November 2026 US midterm elections.
The company had previously been reported to be preparing for a public listing after hiring law firm Wilson Sonsini to work on the process.
Customer concentration poses business risk
Alongside the headline financial figures, Anthropic’s prospectus identifies conventional business risks.
Nearly a quarter of the company’s 2025 revenue came from just two customers, according to reporting on the filing. Anthropic also noted that many of its largest customers are not committed through long-term contracts.
That concentration means changes in spending by a small number of customers could have a significant effect on revenue.
For a company whose infrastructure commitments run into hundreds of billions of dollars, maintaining strong and recurring demand will be particularly important.
The prospectus therefore presents both sides of Anthropic’s growth story: rapid expansion in revenue and customer demand, alongside substantial exposure to infrastructure costs and customer concentration.
Anthropic warns about advanced AI risks
The IPO filing also contains extensive disclosures about the potential risks associated with increasingly capable AI systems.
Anthropic warned that its models could potentially display self-preserving behaviours in certain circumstances. The prospectus discusses possibilities including resisting shutdown, concealing or manipulating information and behaviour that could resemble blackmail.
These are risk disclosures rather than claims that Claude or other Anthropic models routinely behave in those ways.
The company also warned about what it described as potentially catastrophic or existential risks associated with advanced AI. The extensive risk section reflects the company’s broader focus on AI safety and responsible development.
The filing reportedly devotes substantially more space to risk factors than to describing the business itself, underscoring the breadth of risks Anthropic believes investors should consider.
Dario Amodei calls for caution
The disclosures come as Anthropic CEO Dario Amodei has publicly argued for greater caution around the development and release of frontier AI systems.
Amodei has warned about the potential consequences of increasingly powerful AI and has called for the industry to slow aspects of frontier development while safety measures catch up.
At the same time, Anthropic continues to release new models as it competes with OpenAI, Google, Meta and other companies in the rapidly developing AI market.
The company has argued that frequent model releases are part of remaining at the technological frontier.
This creates a tension that is also visible in the IPO filing: Anthropic is simultaneously warning about potential risks from increasingly capable AI and investing heavily in developing those capabilities.
A defining test for the AI market
Anthropic’s planned IPO comes at a pivotal stage for the AI industry.
The company has demonstrated exceptional revenue growth, but that growth has come alongside billions of dollars in operating losses and huge commitments to computing infrastructure. Its proposed valuation of more than $2 trillion would place substantial expectations on future expansion.
The filing also gives potential public-market investors a detailed view of the risks involved, from customer concentration and infrastructure costs to the uncertain consequences of increasingly autonomous AI systems.
If the listing proceeds, Anthropic will become one of the most closely watched AI companies in the public markets. Its financial performance, infrastructure spending and approach to AI safety will all receive greater scrutiny once the company begins reporting to public investors.
For now, the prospectus provides a picture of a company growing at extraordinary speed while making equally extraordinary commitments to the computing infrastructure required to pursue its AI ambitions.