Anthropic draft IPO filing revealed
Archived — this story has rotated out of today’s deck. It is kept here in full.
The gist
Anthropic's draft IPO filing shows first-half 2026 revenue of about $16.3 billion and a net loss of about $81 billion.
The filing also discloses over $600 billion in infrastructure commitments, shaping how investors will price AI risk.
Background
Anthropic is a San Francisco-based artificial intelligence developer preparing a potential IPO that could value it at as much as $2 trillion. The Wall Street Journal reviewed a draft IPO prospectus (a preliminary filing document) that the company has not made public. The numbers are drawing attention because revenue jumped roughly 15-fold from a year earlier while losses and long-term computing commitments also ballooned.
How it unfolded
- Oct 15, 2026Reports of a closed-door mid-October investor briefing said third-quarter revenue of $16.8 billion, up 45% from the second quarter, was discussed; Anthropic had not made its IPO prospectus public.
- Oct 16, 2026MarketWatch, citing a Financial Times report, said the prospectus warned the technology could pose "existential risks to humanity"; the filing was still not public.
- Oct 16, 2026A prediction market treated an Anthropic public-market debut by December 31 as more likely than not, based on its confidential draft S-1 (the IPO registration document) but with no listing date set.
- Oct 20, 2026The Wall Street Journal reported the draft showed first-half 2026 revenue of about $16.3 billion, up from about $1.1 billion a year earlier, and a net loss of about $81 billion, of which roughly $72 billion came from accounting charges tied to convertible notes and other financing instruments.
- Oct 20, 2026The draft showed compute and infrastructure spending of nearly $9 billion in the half, versus about $2.3 billion a year earlier, and infrastructure commitments above $600 billion, including about $450 billion under cloud agreements over the next seven to 10 years. Anthropic declined to comment.
Who’s saying what
- Official
- In the draft filing, Anthropic acknowledged it sought more capacity than hyperscalers would supply, prompting it to pursue third-party data center developments, and flagged risks around model alignment and catastrophic incidents, noting damages from severe AI-related security events could exceed insurance coverage limits.
- Caution
- The prospectus, per a Financial Times report, warned the technology could pose "existential risks to humanity" and flagged potential for AI to display unpredictable and malicious behaviors as risk factors.
- Public
- Mark Levine, the NYC Comptroller, has opened a line of communication with Anthropic to discuss the "existential risks" posed by AI.
Still unverified
Speculation around a November 2026 listing has circulated in multiple reports, but no outlet has confirmed an actual filing date or exchange listing. Several of the most attention-grabbing claims from the investor meeting remain unconfirmed by Anthropic itself, and Anthropic has yet to make its filing public.