Overview
- This week Anthropic confirmed that independent evaluator METR has been given near-employee access to its offices, hardware and internal systems to assess model and misuse risks as part of a public safety pledge.
- Market reports describe a roughly $60 billion, multi-tranche chip financing with yields above 10 percent and Blackstone involvement, which would pay for capital‑intensive compute but raise investor concerns about high borrowing costs.
- The company is actively preparing for a November IPO that some market coverage values as high as $2 trillion, though final terms and pricing remain unconfirmed and subject to change.
- Rivals are responding: OpenAI is accelerating a large private fundraising effort reported at about $30 billion to lock in investor commitments before Anthropic’s planned listing.
- Anthropic is also consulting law firms on how AI agents should negotiate for users and has reported internal shifts in hiring that reduce junior technical roles, a sign that automation is changing workforce demand and company planning.