Anthropic files for IPO with nearly $900 billion in infrastructure commitments

Anthropic has moved closer to a public market debut by filing updated financial documents with the Securities and Exchange Commission. The company revealed its third-quarter performance and long-term spending plans in an amended Form S-1 submission on Monday.

The AI developer generated $33.24 billion in revenue for the nine months ending in September. This figure represents a 1,200 percent jump from the same period in the prior year. Third-quarter revenue alone reached $17 billion. While growth remains strong, the pace has moderated. Revenue expanded by 47 percent in the third quarter, down from 144 percent in the second quarter and 132 percent in the first.

The company posted a net loss of $23.29 billion during the third quarter. For the full nine-month period, the net loss totaled $104 billion. Most of that loss stemmed from changes in the fair value of convertible notes and derivative liabilities. A year earlier, the company reported a loss of $25.1 billion over the same stretch.

The filing details a massive commitment to hardware. Anthropic stated it has entered into purchase commitments and lease agreements with an aggregate value of $895.4 billion. The company noted that substantially all of this spend relates to compute and infrastructure. Compute and infrastructure expenses rose by $14.4 billion over the nine-month period, a 329 percent increase compared to the previous year.

Wall Street observers have expressed concern over the cost of building AI infrastructure. Rising interest rates may make debt unaffordable for data center developers and neoclouds. Quarterly reports from Microsoft, Alphabet, and Amazon recently fueled these worries. Investors are watching whether enough computing capacity will come online to meet Anthropic's demands. The company relies on Amazon, Google, and SpaceX for much of its capacity, while also using neoclouds for expansion.

Dario Amodei co-founded the company with researchers who left OpenAI. OpenAI filed its prospectus in June and targets a 2027 listing. Anthropic's filing includes more than 90 pages of risk factors. The co-founders wrote that they view it as their personal responsibility to help the world understand these risks. They said they aim to prevent or mitigate issues arising from their AI models.

The company will soon begin its roadshow to gauge investor interest. The prospectus outlines risks including political headwinds and the difficulty of balancing safety with commercial goals. Amodei and his team have positioned the firm as more safety-conscious than its rivals.