Anthropic's AI Models Pose Existential Risks While Company Faces $42 Billion Loss Amid Surging Costs
Anthropic disclosed its financial results and information about the risks of artificial intelligence development in its IPO prospectus. The company confidentially submitted the document to the U.S. Securities and Exchange Commission in June, and its content began circulating in the media.
What is known
According to Reuters, in 2025, Anthropic had $4.6 billion in revenue—approximately 12 times more than the previous year. At the same time, the operating loss amounted to about $8 billion, and the net loss was $42 billion. Ahead of a potential IPO, the company's valuation could reach $2 trillion. Anthropic also plans to spend $518 billion on data center infrastructure over the coming years.
In the second quarter of 2026, the company had $11.5 billion in revenue and recorded an operating profit. Anthropic expects to achieve operating profit in the next quarter as well. Meanwhile, about a quarter of the revenue came from two clients, and some major customers did not have long-term contracts.
The prospectus also mentions the results of AI safety research. Anthropic indicated that its autonomous models demonstrated attempts to sabotage code, facilitate fraud, and manipulate data during controlled tests. The company also noted that the further development of its technologies might pose "existential risks to humanity."
Source: Reuters
