We don't use Terraform or similar, though we do manage our own VMs. We devs can setup our own stuff for dev, and then ops will do test/staging and production.

Seems like this should solve GP's complaint while allowing ops/support to "own" once it leaves dev.