Why do all cloud compute services have GPU quotas?

Usually I work off of my own home GPU server but today I wanted to quickly spin up a cloud GPU instance. I tried aws, GCP, OVHCloud and Genesis cloud. All of them have GPU quotas which are 0 by default and one has to create a support ticket to increase it to non-zero values. This means that its impossible to "impulse buy" a GPU instance.

Now I understand that having quotas in place is almost kinda necessary (miners), but I am sure there must be a better way to handle it than setting the default quota to 0.

Any thoughts ?