What happened
Anthropic received investor offers valuing the company at $800 billion, driven by a $30 billion annual revenue run-rate from enterprise adoption of its Claude models. Its new Mythos model, capable of autonomously identifying software vulnerabilities, prompted US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell to convene a closed-door meeting with major bank CEOs, warning of systemic cybersecurity risks. The company also resisted Pentagon use of its models for offensive military purposes.
Why it matters
Anthropic's $800 billion valuation, fuelled by enterprise adoption, signals a market shift prioritising AI infrastructure over consumer models. This focus on enterprise adoption shifts unit economics, a key consideration for procurement teams. The Mythos model's ability to find zero-day vulnerabilities creates a security paradox: a powerful defence tool but a systemic risk if misused. For security architects, this implies assuming agentic workflows are untrusted, which limits deployment flexibility and increases the need for strict internal controls.
