Google claims AI first after SQLite security bug discovered

Google claims one of its AI models is the first of its kind to spot a memory safety vulnerability in the wild – specifically an exploitable stack buffer underflow in SQLite – which was then fixed before the buggy code’s official release.

The…