
Privacy-minded AI users: what are y'all using?
Sup folks, thought I'd ask Signal community this question. I’ve been getting a lot more timid about using frontier models for anything remotely sensitive, especially with some of the recent legal stuff around AI chats, data access, and privilege.
The Snowden/PRISM era already showed how much governments can potentially access through private tech infrastructure, and laws like the CLOUD Act make me pretty uncomfortable about putting my whole inner life into a hosted AI service. The Heppner case also got me thinking about how AI conversations may not have anything close to the protections you’d expect from talking to a lawyer, therapist, doctor, etc.
So I’m curious what folks here are using and how they are thinking about it.
I have seen Confer and Venice as options, the former I am a bit skeptical of, even though Moxie created it, due to the Meta partnership (https://confer.to/blog/2026/03/encrypted-meta/) and the latter seems to be a bit more tailored to the more imagination-oriented crowd ;).
I’m not necessarily trying to go full tinfoil-hat/air-gapped homelab. Just looking for something genuinely useful where I can be a little more intentional about what data I’m handing over.
What are you all using?