That's pretty much true for any nefarious use of AI: if bad actors can use AI for something, the same AI techniques can be used to protect against them.

→ View original post on X — @ylecun