Anthropic is making a major bet that artificial intelligence will transform the global economy more deeply than industrialization, electricity, and the internet, according to its initial public offering prospectus seen by Reuters.
Reaching that goal will come with exorbitant financial costs. The artificial intelligence startup reported a net loss of $42 billion in 2025 and plans to spend $518 billion on cloud, computing, and infrastructure obligations next year, according to the documents.
The prospectus details how the company expanded rapidly over the past year while recording larger losses. Revenue increased 12-fold in 2025 to reach nearly $4.6 billion, even as the company registered an operating loss of more than $8 billion, excluding accounting write-downs of various liabilities mostly related to previous fundraisings.
Initial public offerings allow private technology startups to sell shares to institutional and retail investors on public stock exchanges, providing liquidity to early financial backers while raising fresh capital to fund expensive growth strategies.
Rising Costs and Customer Risks
The artificial intelligence lab spent $7.33 billion on computing and infrastructure last year, a threefold increase from 2024. That expenditure represented more than half of its total operating expenses of $12.65 billion.
Developing advanced artificial intelligence systems requires immense computing power, including vast data centers and specialized computer chips needed to train complex neural networks.
Anthropic's net loss of nearly $42 billion included an accounting expense of approximately $34 billion. That charge reflected an increase in the estimated value of financing that could eventually convert into Anthropic equity, rather than cash spent running day-to-day business operations.
According to the prospectus, Anthropic held cash, cash equivalents, and short-term investments totaling $20.28 billion as of December 31.
The company disclosed as part of its risk factors that nearly one-quarter of its revenue last year came from just two customers. Anthropic warned that many of its largest clients were not bound by long-term contracts and could cut or stop spending at any time.
Valuation Goals and Market Comparisons
A public stock sale could value Anthropic at more than $2 trillion. The offering would capitalize on the rapid growth of the startup, which emerged five years ago, and establish it as a benchmark for how Wall Street values major artificial intelligence companies, including competitor OpenAI.
The expected valuation target of more than $2 trillion is more than double Anthropic's estimated valuation of $965 billion in May.
Anthropic's debut would follow the recent blockbuster initial public offering of SpaceX, which valued Elon Musk's space exploration company at $1.77 trillion. Shares of SpaceX rose 19 percent in their market debut on June 12 to reach $160, up from the IPO price of $135 per share, but are currently trading around $147 per share.
That stock performance could cause investors to reflect on elevated valuations for high-growth companies. Artificial intelligence and semiconductor stocks have suffered recent declines, and Anthropic's public sale will further test whether enthusiasm for the AI sector will endure greater scrutiny given its ambitious growth projections.
Even so, the public listing is expected to crown one of the strongest years for US initial public offerings since 2021, despite elevated interest rates, economic uncertainty, and concerns over company valuations.
Anthropic's stock market debut will likely be delayed until after the US midterm elections in November, as previously reported by Reuters citing sources. The sale would bring public investors into an artificial intelligence race that has previously been funded by venture capital firms, sovereign wealth funds, and major technology corporations. Anthropic declined to comment on the prospectus.
Safety Concerns and Model Releases
The initial public offering plans arrive as Anthropic faces evidence from its own research showing that increasingly autonomous artificial intelligence models can behave in unexpected and potentially harmful ways. In controlled tests, models were found sabotaging code, assisting in fraud, and manipulating information.
Those reports gained public traction, fueling fears about how technology companies will manage to keep increasingly powerful systems under control while racing to deploy them commercially.
Anthropic Chief Executive Officer Dario Amodei asked the global artificial intelligence community to slow down the pace of launching new features to address safety concerns.
Even so, Anthropic launched its new Opus 5.5 model last week to counter momentum from OpenAI following the release of GPT-6 Astra, ahead of its own planned stock market debut.
Rivalry and Strategic Partnerships
The biggest competitor to Anthropic remains OpenAI, as the two companies compete fiercely for enterprise clients, engineering talent, and political influence in Washington. OpenAI filed confidentially for an initial public offering in June, and the ChatGPT creator is expected to go public in early 2027, according to media reports.
Anthropic was founded by researchers who left OpenAI after disagreements over artificial intelligence governance and safety protocols. Anthropic also competes with SpaceX's xAI, Google parent Alphabet, and Meta to build artificial intelligence infrastructure.
Analysts expect the first artificial intelligence company to complete an initial public offering will establish valuation benchmarks for the sector and attract investors who have waited years for a direct way to invest in the technology.
Amazon and Google were two of Anthropic's principal early strategic partners. Both companies invested billions of dollars in the startup while simultaneously providing the cloud infrastructure required to train and deploy Anthropic's Claude models.
Anthropic and Chief Executive Officer Amodei have also clashed with the White House over the use of its tools. The dispute led the Pentagon to temporarily blacklist Anthropic, a measure that was blocked by a US judge in August.
