Post

Physician heal thyself: Anthropic’s new TOS tells users they are explicitly responsible for any actions their agents take "through tools, browsers, or connected systems.”

Have you seen any analysis of this anywhere? I’d love to put this in front of some students.