Anthropic Eyes $750B Compute Spend Ahead of IPO - Briefs Finance
What the filing shows about spending and contracts
A draft document for Anthropic's planned stock market listing, reviewed by the Wall Street Journal, shows that the company plans to make huge investments in computing equipment and services. It has agreed to spend about $450 billion on computing over the next seven to 10 years, and most of these commitments are already fixed. The document also breaks down the amounts planned for each partner: $111 billion with Google's cloud computing business, Alphabet's Google Cloud; $110 billion with Amazon Web Services; and $31 billion with Microsoft.
In a separate report, the Journal says Anthropic may spend as much as $750 billion on additional computing power over the coming years. This shows how much money may be needed to build larger AI systems.
Recent costs and accounting losses
The Wall Street Journal, using information from the draft filing, reports that Anthropic recorded a net loss of $80.7 billion in the first half of 2026. That was much higher than its $13 billion loss during the same period in 2025. The report says $72 billion of the loss in the first half of 2026 came from accounting charges. These charges were linked to changes in the estimated value of convertible notes and other financial instruments the company had issued.
Spending is also rising. According to the Journal, Anthropic spent nearly $9 billion on computing power and infrastructure during the six months ending June 30, 2026. It spent $2.3 billion during the first half of 2025. In the first half of 2026, the company also committed more money to securing computing power than it had spent throughout all of 2025, when its spending reached $7.3 billion.
IPO timing, risks, and recent claims about profits
Anthropic, widely seen as one of the leading companies developing AI, is expected to publicly file paperwork to sell shares on the stock market in the coming weeks. This could lead to one of the biggest initial public offerings (IPOs) so far. The company plans to raise money to buy chips, build data centers, and hire workers to develop new AI models.
Bloomberg News has reported that Anthropic told investors it had reached adjusted profitability in the third quarter and was generating positive free cash flow. The company also reported an 85% gross margin and an adjusted operating margin above 11% as of the third quarter. Bloomberg says the formal process of promoting the IPO to investors could begin as early as the week of November 9. Trading could start before the US Thanksgiving holiday, and the company is expected to go public by the end of the year.
Risks highlighted in the prospectus
The draft prospectus points to AI safety concerns. These include the possibility that AI models might try to protect themselves from being shut down. They might also give responses that resemble "blackmail" in an attempt to stop people from turning them off. The document also warns that political opposition could make it harder to build data centers in the United States. According to the Journal, most of Anthropic's computing capacity is located in the US.
The main point is that Anthropic faces huge long-term spending commitments and high infrastructure costs, even as some measures of its adjusted profitability are improving. For investors, this raises the question of how to balance the company's growth plans against its need for cash as its IPO approaches.