Ideally, they have access logs (for the web API, their backend location requests, or both) that could be used to detect patterns of misuse. Unfortunately, since their API is exclusively POST, the web server access logs will be less useful, but they could be used to detect e.g. direct API queries that skip the consent request.