Anthropic, which reported heavy losses, is seeking a valuation of $2 trillion. The IPO could become a test of how investors price AI’s future potential
Anthropic’s IPO documents warn that AI models could resist shutdowns, conceal information or display behaviour resembling blackmail
As AI models grow more capable, testing their safety becomes harder. Anthropic says models aware of being tested can create significant assessment limits
Anthropic plans $518 billion in cloud and compute capex. Analysts expect AI-related investment to absorb more than $10 trillion over five years
A $2-trillion valuation would put Anthropic among the world’s most valuable listed companies, underscoring the extraordinary expectations surrounding AI
Anthropic’s revenue growth and AI advances are striking. Yet huge losses, vast capital needs and safety risks leave the sustainability of valuations unresolved