Cloud / AI notes: testing the trust boundary of an internal assistant
A lightweight threat model for an assistant using internal data: what is trusted, and which actions need human approval.
taxonomy / category
Cloud / AI
Posts following the same line of inquiry.