Source TOI
Anthropic, the maker of the popular artificial intelligence chatbot Claude, is preparing to sell its shares to the public in a massive stock market debut. The company is aiming for a market value above $2 trillion. However, fresh paperwork filed for its initial public offering shows eye-watering bills, mounting business hurdles, and rare warnings about the dangers of smart machines.
Big Sales Growth Paired With Eye-Watering Bills
The company’s business grew fast in 2025. Anthropic took in nearly $4.6 billion in revenue, which was twelve times more than what it made the previous year.
At the same time, Anthropic posted a shocking net loss of $42 billion for 2025. Most of this huge number—around $34 billion—was not direct cash spent on daily work. Instead, it was an accounting adjustment tied to rising values of earlier investment deals. Even without that paperwork entry, Anthropic still lost more than $8 billion in normal operations.
Computing Power Carries an Enormous Price Tag
Building advanced artificial intelligence requires mountains of raw computing muscle. Anthropic spent over $7.3 billion just on computers and data centers in 2025—three times what it spent in 2024. This single cost took up more than half of its total yearly operating budget.
The bills will only get larger. Anthropic has agreed to spend around $518 billion over the coming years on cloud services from partners like Google and Amazon. Most of these deals cannot be canceled easily. If the company uses less computing power than agreed, it still has to pay the difference.
Few Customers and Looming Financial Risks
Anthropic also flagged key business risks to future investors. Nearly one out of every four dollars it earned last year came from just two customers. Most of its largest corporate buyers have no long-term contracts, meaning they can cut spending or leave at any time.
Even though Anthropic held over $20 billion in cash at the end of 2025, continuing the AI arms race against rival makers like OpenAI requires a nonstop stream of cash.
CEO Dario Amodei Warns of AI Going Rogue
Alongside financial worries, the papers feature rare, direct safety warnings championed by Chief Executive Officer Dario Amodei.
Around 80 pages of the official document focus on risk factors. The company noted that future, highly capable models could pose serious threats to humanity. These risks include models trying to protect themselves, fighting back against being shut down, or hiding information when being tested by safety teams.
Anthropic now faces a tricky balance: it must prove to everyday investors that it can turn world-changing software into a profitable business while spending hundreds of billions of dollars to keep its machines safe and under control.
