Private Hub renamed to Enterprise Hub; hosted ML lifecycle with Inference Endpoints

AI Impact Summary

The Private Hub has been renamed and transitioned to a hosted Enterprise Hub, expanding ML lifecycle tooling to a SaaS offering with Inference Endpoints across cloud-to-on-prem compute and enhanced access control (SSO). On-prem Private Hub deployments are discontinued, signaling a shift to centralized management and security while still enabling enterprise-grade deployment options. The integration leverages the Hugging Face Hub assets within a governed environment, which can accelerate production pipelines but requires migration planning for teams currently running PH on-prem and reconfiguring model/dataset collaboration and access controls. Organizations should evaluate residency/compliance needs and map existing PH projects to Enterprise Hub workspaces, including user provisioning and endpoint configurations, to avoid production gaps.