VPC DA regularly fails with Error: AccessDenied: Access Denied 403

[ocofaigh]

The VPC DA is failing tests regularly with errors like this:

 2025/05/24 23:31:45 Terraform apply | Error: AccessDenied: Access Denied
 2025/05/24 23:31:45 Terraform apply | 	status code: 403, request id: 03831eec-db86-4bf6-8688-f6bc0e722b41, host id: 
 2025/05/24 23:31:45 Terraform apply | 
 2025/05/24 23:31:45 Terraform apply |   with module.cos_buckets[0].module.buckets["landing-dbce-flow-logs-bucket"].ibm_cos_bucket.cos_bucket[0],
 2025/05/24 23:31:45 Terraform apply |   on .terraform/modules/cos_buckets/main.tf line 122, in resource "ibm_cos_bucket" "cos_bucket":
 2025/05/24 23:31:45 Terraform apply |  122: resource "ibm_cos_bucket" "cos_bucket" {
 2025/05/24 23:31:45 Terraform apply | 
 2025/05/24 23:31:45 Terraform apply | ---
 2025/05/24 23:31:45 Terraform apply | id: terraform-d2853781
 2025/05/24 23:31:45 Terraform apply | summary: "AccessDenied: Access Denied\n\tstatus code: 403, request id:
 2025/05/24 23:31:45 Terraform apply | 03831eec-db86-4bf6-8688-f6bc0e722b41,
 2025/05/24 23:31:45 Terraform apply |   host id: "
 2025/05/24 23:31:45 Terraform apply | severity: error
 2025/05/24 23:31:45 Terraform apply | resource: ibm_cos_bucket
 2025/05/24 23:31:45 Terraform apply | operation: create
 2025/05/24 23:31:45 Terraform apply | component:
 2025/05/24 23:31:45 Terraform apply |   name: github.com/IBM-Cloud/terraform-provider-ibm
 2025/05/24 23:31:45 Terraform apply |   version: 1.78.2
 2025/05/24 23:31:45 Terraform apply | ---

It is intermittent, but frequent. Is there some timing issue? Is there CBRs getting enforced in our account by some test? Is there something in our code we can do to help prevent this?