Dario Amodei, the CEO of Anthropic, spoke at the AI Impact Summit in New Delhi on February 19, 2026, revealing the company’s plans to go public. This recent IPO filing illustrates the ambitious vision Anthropic has for artificial intelligence, likening its transformative potential to historical milestones like industrialization, electricity, and the internet.
According to their prospectus, Anthropic reported a staggering net loss of nearly $42 billion for 2025 and anticipates assuming future obligations totaling approximately $518 billion related to cloud computing, infrastructure, and computing capabilities. In contrast, the company's revenue soared to nearly $4.6 billion in 2025, signifying an impressive twelvefold increase from the previous year's revenue, highlighting the explosive growth that underpinned its decision to enter the public market.
The IPO prospectus provides insight into the high costs associated with Anthropic's aspirations, revealing that the company spent $7.33 billion on computing resources and infrastructure in 2025, tripling its expenditures from 2024. This expenditure accounted for more than half of Anthropic's total operating expenses of $12.65 billion. Notably, over $34 billion of the reported loss stemmed from an accounting adjustment related to financing that could convert into shares rather than from ordinary business operations.
As of the end of 2025, Anthropic had approximately $20.28 billion available in cash, cash equivalents, and short-term investments. Reports suggest that the company could aim for a valuation exceeding $2 trillion, significantly higher than the roughly $965 billion estimated during its previous private financing round. The ultimate valuation will hinge on investor interest and the finalized offering details.
Anthropic’s growth trajectory is evident in its San Francisco presence, as the company has leased virtually the entire 25-story building at 300 Howard St.—around 480,000 square feet—as its new headquarters. Additionally, it has expanded into more space at 400 and 500 Howard St., solidifying its position near the Salesforce Tower in the South of Market area.
As a public benefit corporation, Anthropic seeks to harmonize its commercial objectives with its mission to responsibly advance AI for the betterment of humanity. However, the IPO filing also reveals potential risks: nearly a quarter of the company’s revenue originated from two unidentified clients last year, and many key customers lack long-term contracts, which could affect future revenue if they choose to reduce their AI spending.
The company’s operations are heavily reliant on extensive computing power; as its AI models enhance in capability, the demand for computing capacity will escalate.
Anthropic’s move into public markets is coming at a time of rapid advancements in AI, with competitors like OpenAI, Google, Meta, and Elon Musk's xAI also vying for market share and resources. Anthropic's IPO may serve as a crucial gauge for how Wall Street values an AI enterprise that is balancing rapid revenue growth with significant infrastructure investments.
Additionally, the company is dealing with an intricate relationship with the Trump administration. CEO Dario Amodei had a private meeting with President Trump recently after tensions regarding AI policy and national security issues. A few days before this meeting, a federal appeals court upheld a Pentagon decision excluding Anthropic’s Claude models from its supply chain, reasoning that the department had valid grounds for its decision.
Anthropic has raised concerns over the potential uses of its technology, particularly in lethal autonomous weapons and mass surveillance, while the Pentagon has sought broader access to its models. Ultimately, the court ruling did not conclude that the Pentagon’s actions were influenced by Anthropic’s push for AI regulation.
Amodei is scheduled to return to the White House on Tuesday, this time alongside other prominent figures from the AI sector.
