Paper 2026/1443

SENTRA:Privacy-Preserving Training in Outsourced Cloud Environments

Abstract

Training machine learning models in untrusted clouds requires strong guarantees of confidentiality, integrity, and correctness, while remaining scalable and resilient to node churn. These challenges are further amplified in emerging agentic AI systems, where autonomous and distributed learning components require trustworthy coordination and secure state management across heterogeneous cloud environments. Existing Trusted Execution Environments (TEEs) lack scalability and remain vulnerable to side-channel attacks for large workloads, while pure secure multi-party computation (MPC) approaches incur prohibitive overhead in practice. SENTRA (Secure ENclave-based TRaining Architecture) addresses these challenges through a hybrid architecture that combines TEEs, secret sharing, and communication-efficient MPC with system level mechanisms that secure the entire training lifecycle. SENTRA introduces a scalable collective attestation protocol that verifies all participating enclaves and enforces hardware exclusivity before any node may store or process secret shares. Training data and model parameters are stored as secret shares in a versioned enclave-backed key–value store (KVS), providing rollback protection and consistent state under adversarial conditions. SENTRA further supports dynamic, fault tolerant membership through Dynamic Proactive Secret Sharing (DPSS)-based resharing, safe packed-MPC computation under degree bounds, and adaptive handling of node failures. Evaluation of a prototype implementation shows that SENTRA achieves up to 8.89 samples/s throughput and 1.29× faster training than the CrypTen baseline in software-only mode. In hardware-enclave mode, SENTRA incurs only an 8.3% performance overhead while providing memory-isolated confidentiality, fault-tolerant membership management, rollback protection, and recovery from node failures in approximately 8 seconds.