Abstract: In general, this disclosure describes a multi-zone secure AI exchange. The multi-zone secure AI exchange may be implemented in a multi-cloud, multi-data center environment, where each zone may be in a different cloud or data center. The multi-zone secure AI exchange may include a data repository, a data exchange, and shared services. The data repository may be configured to store algorithms and datasets, each having a respective owning user. The data exchange may receive datasets and algorithms from the data repository, and may perform the algorithms to produce output data. Each of the data repository, data exchange, and shared services may have a different level of security. The data repository may implement the highest level of security, allowing the owner user, and only the owning user, to control how their data and algorithms move in and out of the data repository, or are changed while in the data repository.

Abstract: In general, this disclosure describes a multi-zone secure AI exchange. The multi-zone secure AI exchange may be implemented in a multi-cloud, multi-data center environment, where each zone may be in a different cloud or data center. The multi-zone secure AI exchange may include a data repository, a data exchange, and shared services. The data repository may be configured to store algorithms and datasets, each having a respective owning user. The data exchange may receive datasets and algorithms from the data repository, and may perform the algorithms to produce output data. Each of the data repository, data exchange, and shared services may have a different level of security. The data repository may implement the highest level of security, allowing the owner user, and only the owning user, to control how their data and algorithms move in and out of the data repository, or are changed while in the data repository.

Abstract: In general, this disclosure describes a system for securely training artificial intelligence (AI) models. The system may include communication circuitry for receiving from providers data sets, machine learning algorithms, and AI models. The data sets, machine learning algorithms, and AI models may be placed in a secure sandbox for access by a user. The user may securely train the AI models using the data sets in the secure sandbox. The secure sandbox may record AI model metadata associated with transactions involving the AI models and send the AI model metadata to a model registry. The model registry may attest to the lineage of the AI models.

Abstract: In some examples, an application programming interface (API) and services engine is configured to determine, in response to receiving a service request from an application, an application identifier for the service request based at least on an authorization token included in the service request, wherein the application identifier identifies an application that issued the service request; select, based at least on the application identifier, an application data store uniquely associated with the application that issued the service request; and process the service request using data stored by the selected application data store.