Abstract: System and methods use a network service registry to differentiate risks associated with use of respective network services, e.g., artificial intelligence (AI) and/or machine learning (ML) network services that may pose risks to use in an enterprise context based for example upon training, data retention, and/or content moderation policies of the respective network services. Upon a request (e.g., outbound network traffic) to an intended network service, systems and methods may intercept the request, reference the request to the service registry to identify risk identifiers of the intended network service, and select an approved network service(s) for the request (e.g., the same intended network service, and/or a different service). Thus, the systems and methods may mitigate or prevent potentially hazardous use of AI and/or ML network services in an enterprise.