I want to ask if it is possible to detect the abuse of a service in the cloud. How can the SaaS provider be sure that the client does not use the service for criminal acts? For example how can Google be sure that someone does not store pedophile pictures in Google Drive?
It is possible to programmatically discover known pedophile images by hashing the user's files and checking for matches. Governments request data from Google for criminal investigations so such efforts are conducted in a legal context. Google provides details on the quantity of those requests and references to the legal aspects here.