With as much as AI like chatGPT are used to aid in programming is anyone working on developing AI as an iterative attack surface to probe networks?

There are some obvious ethical concerns, but I can see a future with AI red and blue team actors playing a substantial role in network security.