Jamie, Something I don't think you understand is that no enterprise cares if they are using IPv4. They just care that their network works. They aren't religiously attracted to IPv4. If IPv6 saves them money they will switch. Currently they probably have an IPv4 only network. In 2026 that's technical debt, but the business can decide when it makes sense to them to clear up that technical debt. They probably don't have routers old enough to not support IPv6 at this point, and if they do, they can roll IPv6 deployment into their router upgrades which are also overdue. As long as they carry that technical debt, they will also be carrying the real costs of running an IPv4-only network: - Leasing / owning IPv4 blocks - Leasing IPv4s for cloud resources - Engineering time dealing with issues related to doing NAT across business domains - Engineering time dealing with issues related to IPs not being unique (i.e. re-used 10/8 space) So, how much is the enterprise spending dealing with these issues? Rolling out dual stack (not even IPv6-only!) means internal resources are directly addressable and NAT becomes something you only do for your internet-bound v4-only traffic (where you don't really care if every location uses the same 192.168.1.0/24, it's just for internet access), not within the corporate network. No need to build a new stack, no need to have a ton of translators everywhere, no need to replace every single NIC. Rolling out IPv6-only (i.e. 464xlat, or dns64) means you get to completely remove the engineering time you are spending dealing with IPv4-related issues, not just minimize it. I'm not sure what issues you think you are solving with all the other junk you've thrown into the protocol (DHCP granting everything, OAuth, ..) but those are also very solved issues in enterprise networks that don't involve replacing hardware with validated firmware. IP addresses are not proof of identity, not in enterprise networks, not anywhere, they are a routing location of where a client is logically located within the network. Treating them as a proof of identity is a failure of your security implementation, not a fault in the protocol. This goes for both IPV4 and IPV6. Andrew