These are fair points and I don’t think the app resolves them - it’s a tool, and tools can be misused regardless of design. The most I can say is what I aimed

These are fair points and I don’t think the app resolves them - it’s a tool, and tools can be misused regardless of design. The most I can say is what I aimed for: showing the source text in full rather than just an AI-generated verdict, so the user is reading Wilhelm and Jung’s actual words, not a black box. Whether that’s enough safeguard against the deeper concern you’re raising, I don’t know.