Fortanix Adds Confidential Containers Support to Confidential AI
Fortanix announced support for Confidential Containers in its Confidential AI platform, connecting Kubernetes workload attestation with policy-controlled encrypted-key issuance. According to the company, workloads that pass validation can decrypt protected models and sensitive data inside hardware-backed trusted execution environments. The protections still depend on trusting guest software, configuring attestation correctly, and enabling appropriate GPU verification.
The integration gives teams a vendor-provided path to run sensitive AI models and data across shared infrastructure, while also making the trust and configuration boundaries of that protection explicit.

Fortanix 宣布"在其 Confidential AI 平台中支持 Confidential Containers。该集成打通了 Kubernetes 工作负载认证与加密密钥发放流程。因此,符合安全策略的工作负载可以在基于硬件的可信执行环境(TEE)内解密模型和敏感数据。
Confidential Containers"(CoCo)是云原生计算基金会(CNCF)的一个项目,它能够将 Kubernetes Pod 运行在由基于硬件的可信执行环境(TEE)作为底层支撑的机密虚拟机内部 。与普通容器不同——隔离性主要依赖于宿主操作系统——机密虚拟机利用硬件保护机制对内存进行加密,并在工作负载运行期间限制对其访问。这有助于保护“使用中的数据”(而不仅仅是“静止数据”或“传输中的数据”),并允许团队在共享基础设施上运行敏感工作负载。不过,这些保护措施仍然需要信任客户机软件,并正确地配置认证和密钥发放机制。
#Fortanix#Confidential AI#Confidential Containers#Confidential Computing Manager#Data Security Manager#Trustee#Azure Secure Key Release#Intel TDX#AMD SEV-SNP#NVIDIA