I mean we’ll see, in general stuff like finding vulnerabilities in large code bases seems like a good fit for this tech. All it’s doing is making statistical inferences based on training, and this can help spot problems that would be hard to track down by hand.