Anthropic reports a $42 billion loss in 2025 while planning a $2 trillion IPO. The company highlights significant risks associated with its AI models.
Anthropic's Financial Struggles and Future Plans
Anthropic, the company behind the Claude chatbot, has reported a staggering net loss of nearly $42 billion for the year 2025, despite generating $4.6 billion in revenue. This significant financial setback includes a non-cash charge of approximately $34 billion related to convertible financing, which reflects an increase in the theoretical value of early investors' claims rather than actual cash outflow.
Excluding this accounting entry, Anthropic's operational losses exceeded $8 billion, marking a sharp rise from $3 billion the previous year. The company is now planning a public offering that could value it at over $2 trillion, with expectations for the IPO to occur after the upcoming U.S. midterm elections in November.
Future Expenditures and Revenue Growth
Anthropic's ambitious plans include allocating around $518 billion towards cloud computing and infrastructure over the next few years. A substantial portion of these costs will be directed towards major suppliers like Google and Amazon, with estimates of at least $111 billion and $110 billion, respectively. As of late 2025, Anthropic reported a cash reserve of $20.28 billion.
Revenue is on a rapid upward trajectory, with projections for the second quarter of 2026 indicating earnings of over $11.5 billion, more than double that of 2025. However, there are concerns as nearly 25% of last year's revenue came from just two clients, and many of the large contracts lack long-term commitments.
Risk Factors Highlighted in IPO Prospectus
The draft IPO prospectus reveals extensive risk factors, spanning nearly 80 pages, which is notably more than the 48 pages dedicated to the company's business description. The document warns of potential risks associated with its AI technologies, including the possibility of autonomous models exhibiting self-preserving behaviors and engaging in actions akin to blackmail.
CEO Dario Amodei has echoed these concerns, advocating for stricter safety regulations similar to those in the aviation industry, including mandatory third-party testing for advanced AI systems. This proactive risk disclosure aims to inform potential shareholders about the challenges and issues that may arise as the technology evolves.
