Anthropic's S-1 is the most honest document the AI boom has produced

Revenue hit $16.3 billion in the first half of 2026, up from $1.2 billion a year earlier, while the net loss came to $80.7 billion. That headline loss is mostly noise: roughly $72.4 billion of it came from fair-value accounting on convertible notes, and the actual operating loss was $693 million, down from $2.1 billion a year ago. The real number to watch is compute spending, which hit nearly $9 billion in six months against $7.3 billion for all of 2025, alongside roughly $800 billion in disclosed data-center commitments cited in the filing. Anthropic wants to list on Nasdaq as ANT with Goldman Sachs and Morgan Stanley leading more than 20 banks, at a reported target valuation of $1.5 trillion to $2 trillion, which would make it bigger than the SpaceX IPO this summer. The filing devotes over 90 pages to risk factors including military use, autonomous AI behaviour and catastrophic misuse, which is either admirable candour or the sort of disclosure that makes plaintiffs' lawyers very happy five years from now. Founders retain 50.1% of voting power after listing, so public shareholders are buying growth without buying control.