"Does 'runs locally, not the cloud' actually build trust, or is that just founder cope?"

Genuine question for anyone who's thought about AI agents with deep access to your data — email, files, calendar, that kind of thing.

Does it actually change your trust calculus if the AI runs entirely on your own machine instead of a cloud provider's servers — or is "local-first" more of a founder talking point than something people actually weigh when deciding whether to give a tool that kind of access?

Asking because I've spent the last few months building a personal AI assistant that runs fully local, specifically on the assumption that people would trust it more with sensitive access. But I don't actually have real evidence for that — nobody's told me "local" was the thing that changed their mind.

So: if you've used, or considered and rejected, an AI tool that wanted deep access to your inbox/files — what actually made you comfortable, or not? Was where it runs ever part of that decision, or does it really come down to something else (company reputation, specific permission scoping, data retention policy)?

Genuinely curious what this community thinks — not trying to sell anything here, just trying to find out if I've been building around a real insight or a nice-sounding assumption.