Is this what zero trust is supposed to protect against? It's weird to let everyone have free reign over critical infrastructure and data once you gain access to the network.

Are there any network security that uses AI to automatically detect and flag unusual behavior? For example, a regular developer doesn't need to port scan the entire network, the receptionist's computer doesn't need to log in to admin resources, support personnel doesn't need to access source code, etc. This should be possible to flag and let a human verify.