Anthropic S-1: revenue $41M to $16.3B in half a year, ~900% NRR, and AI-attacker warnings
colinmcnamara · x · 2026-10-26
Anthropic's S-1 filing shows revenue growing from $41M in 2023 to $16.3B in H1 2026, 1,900+ customers spending $1M+/year, and 900% net revenue retention; Opus 5→5.5 cut cost per task 40%. Reading it as a security builder: the prospectus admits attackers already use AI to "discover vulnerabilities, generate malware, and evade detection tools," and warns each Claude model helps build the next, with capabilities potentially outpacing safety measures — citing Mythos and Project Glasswing as limited releases. A "two loops, one clock" framing contrasts the offense loop with a defense loop built on defenders' own environments.
Related event: Anthropic Files S-1, Kicking Off IPO Under Ticker ANT on Nasdaq(71 posts)→
More from Venture
- Anthropic's $80.7B "loss" is mostly paper: fixed-price share rights to Google/Amazon at $259 vs $589 — rohanpaul_ai · 2026-10-27
- Bullish on Anthropic's IPO thanks to Opus 5.5 — but wary of Dario's CNBC mouth — CtrlAltDwayne · 2026-10-27
- 14 custom Lego orders shipped, powered by AI-generated brick designs — noahsolomon · 2026-10-27
- Anthropic officially files for IPO under ticker symbol "ANT" — connoraxiotes · 2026-10-27
- Boaz Barak endorses Zitron's old take: Anthropic 'is not worth $60 billion' — inductionheads · 2026-10-27
- Debate over 'China's AI bubble burst': stocks fell 40% but companies still shipping — TinfoilTricorn · 2026-10-27