PPrecision AI OS·9h agoWhat Anthropic's amended S-1 adds on agent oversight and liabilityAmendment No. 1 keeps the reviewer warnings, adds a line about human approval inside Anthropic's own research, and spells out an open liability question3 min read·precisionaios.hashnode.dev