Anthropic disclosed a $517 billion compute commitment covering 14.8 gigawatts of capacity over the 11 months ending August 2026, according to a report in The Information. The figure far exceeds an earlier investor disclosure of roughly $180 billion in compute spend projected through 2029.
The AI startup said its revenue run‑rate surged from about $9 billion to $65 billion within seven months, based on investor briefings reported by the Financial Times. Anthropic also told investors it expects a second consecutive quarter of adjusted operating profit and that inference‑stage gross margins topped 80 percent before distribution costs.
Anthropic is preparing for a Nasdaq listing with a target valuation of roughly $2 trillion, Reuters cited. The valuation follows a private‑funding round that placed the company at $965 billion. The firm confidentially filed its IPO prospectus with the U.S. Securities and Exchange Commission in June 2026 and expects to list before the November 2026 U.S. midterm elections.
Nvidia is reported to be in talks to invest up to $10 billion as an anchor investor in the anticipated offering. Nvidia CEO Jensen Huang has projected global AI‑infrastructure spending of $3 trillion to $4 trillion by 2030 and expects about 70 percent revenue growth in fiscal 2027, citing a supply ceiling rather than demand constraints.
In a separate statement, OpenAI CEO Sam Altman said the company will not pursue a public listing in 2026, describing the timing as “ill‑advised.”
Anthropic’s chief executive Dario Amodei used an essay published on a Saturday to call on AI labs to “slow the pace at which we improve the capabilities of AI models,” proposing third‑party evaluators, industry‑wide safety standards and global coordination. He added that “progress will still seem fast, and we must make wise use of the time we gain.”
Analyst Gil Luria of D.A. Davidson noted that investors may not view the slowdown call as a negative, while Bernstein analysts led by Stacy Rasgon questioned whether concerns about China are influencing the proposal and warned that “it is far too late to put the AI genie back in the bottle.”
