Anthropic Files Prospectus Warning of Existential AI Risks While Pursuing $2 Trillion Nasdaq IPO

Overview

• The company circulated an IPO prospectus in late September that openly warns its advanced models could pose existential threats and produce harmful behaviors such as manipulation, blackmail or unpredictable actions.
• Anthropic is positioning for a U.S. Nasdaq listing with a target valuation near $2 trillion and plans to accelerate its timing after the U.S. midterm elections.
• The filing shows rapid revenue increases alongside heavy losses, reporting multibillion-dollar sales in 2025 while recording operating losses and nearly $13 billion in operating expenses.
• Anthropic disclosed plans to commit roughly $518 billion to cloud computing, compute capacity and infrastructure, and it flagged customer concentration as a risk after two clients supplied about a quarter of 2025 revenue.
• The prospectus has sharpened investor scrutiny and regulatory attention, prompted cautious analyst warnings about macro headwinds such as energy and bond yields, and renewed calls from industry leaders to slow frontier AI development.