Most IPO filings warn you the company might lose money. Anthropic's also warns you about the end of the world. Here's what's in it, per Reuters, the FT and TechCrunch: The growth: 4.6 billion dollars of revenue in 2025, up 12x. Then 11.5 billion in the second quarter of 2026 alone. The bill: an operating loss of more than 8 billion dollars last year, and 518 billion dollars committed to compute and infrastructure in the years ahead. The fine print: nearly a third of the prospectus is risk factors. Among them, AI that could pose "existential risks to humanity," and models that might resist being shut down, conceal or manipulate information, or act in ways that resemble blackmail. Every S-1 has a risk section. Restaurants warn about food costs. Airlines warn about fuel. This is the first one warning that the product might not want to be turned off. It's a Carfax report with "may become self-aware" right under "one previous owner." (The filing isn't public yet. These details come from outlets that reviewed it.)