Fortanix extends Confidential AI to Kubernetes workloads

Fortanix Inc. has expanded its Confidential AI platform to support the Cloud Native Computing Foundation’s Confidential Containers project. This update allows enterprises to run artificial intelligence workloads on Kubernetes within hardware-based trusted execution environments.

The move addresses security vulnerabilities that occur during AI inference, a stage where model weights and input data must be decrypted in processor memory, potentially exposing them to privileged administrators or cloud operators. According to a survey of 300 AI leaders at large U.S. companies conducted by Atomik Research for Fortanix, 93% of respondents believe traditional encryption fails to protect data during AI processing, and 41% have blocked deployments due to security concerns.

Fortanix is developing open-source plugins for the Confidential Containers project to verify container authenticity before granting access to encryption keys. The process utilizes the Fortanix Confidential Computing Manager to attest both the CPU and GPU, ensuring full coverage for inference jobs. The company’s Data Security Manager only releases keys if the results meet specific customer security policies.