Comment by unethical_ban

I think in time (a few years) the gap will close a bit. As AI is used to harden software, it will be harder to break. Software not validated by AI will continue to be vulnerable to it.