100 nasty bugs in the code
100 bugs in the code
Take one down
Patch it around
-127 nasty bugs in the codeHow is having Copilot breach trust and privacy an “advisory”? Am I missing something?
The bug is fixable, but the underlying tension—giving AI tools enough permissions to help while respecting confidentiality boundaries—will keep surfacing in different forms as these tools become more capable.
We're essentially retrofitting permission models designed for human users onto AI agents that operate very differently.
tablets•2h ago