Fortanix brings Confidential AI to Kubernetes containers

Fortanix has extended its Confidential AI platform to support the CNCF's Confidential Containers project, enabling enterprises to run AI workloads on Kubernetes inside hardware-based trusted execution environments. The update addresses exposure during inference, when model weights and data must be decrypted in processor memory. Fortanix's open-source plugins verify container integrity before any encryption key is released, with attestation covering both CPU and GPU together. Deployment into a confidential environment requires one click on supported hardware. The platform supports on-premises and air-gapped deployments, and is available now. AMD, Intel, CrowdStrike, and Nvidia are among supporting partners.