Open Medical Interference (OMI) is a project in the second funding phase of the Medical Informatics Initiative (MII) consortia and aims at enabling the usage of remote artificial intelligence (AI) services. As AI methods rapidly advance, time-consuming and repetitive tasks in medicine can be adopted using such services. However, the deployment of multiple AI services from different vendors puts high demand on local IT infrastructures and IT staff. Likewise, the enormous cost of local infrastructure to maintain a wide range of AI models designed for rare cases, should be in consideration. Thus, a cloud-based service, as intended in the OMI use case, offers to achieve semantically interoperable peer-to-peer exchange of multimodal healthcare data and remote AI inference, open protocol and specified data formats will be adopted. OMI components include a gateway server to connect AI services to the MII data sharing framework, a client to enable Data Integration Centres (DIC) and data management service providers to access OMI gateway servers, and a service registry to discover and check the status of connected AI services. OMI targets at extending the MII core data set to include medical imaging in terms of Fast Healthcare Interoperability Resources (FHIR) description of medical image datasets and AI derived metadata. Furthermore, additional DIC components that include data protection measures and enable structured and user-defined DICOMwebTM based access, transport of DICOM data and DICOM de-identification services will be implemented.