GrapheneOS on Nostr: We didn't say that at all. We said we've started using AI models for extra passes of ...

We didn't say that at all. We said we've started using AI models for extra passes of code review for our own code and that it consistently finds real issues missed by human code review.

We could use it to find a large amount of serious bugs in many of the upstream open source projects we depend on lack resources to identify and resolve a large number of upstream security bugs ourselves. There's already a massive effort doing that for the Linux kernel and many other projects without us.