There needs to be something. I'm curious what we will come up with. Some kind of sentinel that sits on the network and does behavioural and threat analysis of all network traffic and prevents damage before it can happen perhaps?

When I hear something like "network police", I start drawing parallels to the real world. In the physical world, you generally don't have anonymous people running around (or at least most cases you have a way of linking any actions/crimes committed to a real, permanent identity), disguises aren't allowed, and trespassing is not permitted.

So the Internet could be like this if it was more regulated. Anonymous traffic could be prohibited...no more TOR nodes, no hands-off proxying of traffic, no "it's an open access point, I totally don't know who was creating that torrent traffic".

Would these sorts of laws be accepted, or would they simply result in more attempts to anonymize traffic?

I imagine that this is sort of what things are like in more authoritarian places like China. Is it effective there?

Such sentinel devices have been available for years. The trouble is they can't reliably identify attempts to exploit zero-days because the patterns are unknown.

Sure we do. Anti-money laundering regulation is basically just the government outsourcing the front-line policing of financial transactions to banks and businesses.

Anti-money laundering regulation requires you to obtain and log information about the customer, not to recognize money laundering and definitely not to solve it (except in very special cases applicable to mega-corporations).

I’ve been the AML Officer for several large financial institutions and I can assure you they absolutely do need to recognize money laundering. They’re required to have systems, policies, and procedures in place to identify, prevent, and detect money laundering.

Keyword "large". Yes, we require large companies to do much more than small ones. But we require even the small ones to deal with network breaches perfectly.