According to its initial public offering (IPO) prospectus, Anthropic is making a major bet that artificial intelligence will transform the global economy more profoundly than industrialization, electricity, and the internet.
However, achieving this goal will require colossal expenditures. In 2025, Anthropic recorded a net loss of $42 billion and plans to spend $518 billion on cloud services, computing power, and infrastructure needs in the coming years, as stated in the prospectus.
Documents first published by Reuters detail the company's sharp growth over the past year, accompanied by increasing losses. Revenue grew twelvefold in 2025, reaching nearly $4.6 billion, even though operating losses exceeded $8 billion, excluding write-offs of various liabilities mainly related to previous fundraising.
The IPO, which Anthropic may value at over $2 trillion, will allow it to capitalize on the rapid ascent of the five-year-old AI lab and establish a benchmark for valuing leading AI companies on Wall Street, including competitor OpenAI.
These plans emerge against a backdrop where Anthropic faces data from its own research indicating that increasingly autonomous AI models can behave unpredictably and potentially dangerously, including code sabotage, aiding fraud, and manipulating information in controlled tests.
These reports have entered the wider public sphere, raising concerns about how companies can control increasingly powerful systems as they are commercially deployed. Anthropic CEO Dario Amodei has called on the global AI community to slow the pace of releasing new capabilities to address these issues. Nevertheless, Anthropic released its new Opus 5.5 model last week to counter the momentum of OpenAI following the launch of GPT-6 Astra, just before its anticipated IPO.
The AI Race
The AI lab spent $7.33 billion on computing and infrastructure last year, a threefold increase compared to 2024 and accounting for more than half of total operating expenses of $12.65 billion. Anthropic's IPO, according to Reuters, is likely to be postponed until after the mid-November US elections. The sale will attract public investors into a race that has so far been funded by venture capital funds, sovereign wealth funds, and large technology companies. Anthropic declined to comment on this matter.
Anthropic's target public market valuation is projected to be more than double its own valuation of $965 billion set in May. The net loss of nearly $42 billion included an accounting charge of approximately $34 billion, reflecting an increase in the assumed value of financing that may eventually convert into Anthropic shares rather than money spent by the company on business operations.
Anthropic reported that almost a quarter of its revenue last year came from two clients, and in the risk factors section warned that many of its largest customers are not tied to long-term contracts and may reduce or cease spending.
According to the prospectus, as of December 31, Anthropic held cash, cash equivalents, and short-term investments totaling $20.28 billion. Anthropic's debut will follow the recent high-profile IPO of SpaceX, which valued Elon Musk's company at $1.77 trillion. SpaceX shares rose 19% to $160 at its June 12 debut but currently trade around $147—still above the IPO price of $135.
This dynamic may cause doubt among investors who are assessing still high valuations for high-growth companies. Stocks in the AI and semiconductor sectors have recently fallen, and Anthropic's offering will further test whether the enthusiasm for AI trading can withstand closer scrutiny, given its ambitious growth forecasts.
Nevertheless, the listing could secure one of the strongest years for US IPOs since 2021, despite higher interest rates, economic uncertainty, and valuation concerns. Anthropic's main rival remains OpenAI, as both firms fiercely compete for business clients, talent, and influence in Washington. OpenAI confidentially filed for an IPO in June, and media reports suggest the ChatGPT developer may go public by early 2027.
Anthropic was founded by researchers who left their competitor over disagreements regarding AI governance and safety. The company released its first large language model in March 2023, directly competing with OpenAI. It also competes with SpaceX's xAI, Alphabet's Google, and Meta in building AI infrastructure.
Analysts expect the first company to go public to set industry benchmarks for AI and attract investors who have waited years for a clear way to bet on the technology.
Billions of Dollars
Amazon and Google were among Anthropic's early strategic partners. Both companies invested billions in it while providing the cloud infrastructure necessary for training and deploying Claude models. Anthropic and Amodei clashed with the White House over the use of its tools, leading to the company being placed on the Pentagon's blacklist. A US judge blocked this designation in August, but a federal appeals court in Washington upheld a separate designation of Anthropic as a Pentagon supply chain risk, according to CNBC.