A hardware-based security paradigm that protects data in use by isolating computation within trusted execution environments (TEEs) backed by processor security extensions (Intel SGX, AMD SEV, ARM TrustZone). It extends encryption from data at rest and in transit to data actively being processed, preventing access even by privileged software, hypervisors, or cloud providers. Key AI applications include secure model training, private inference, and TEE-protected federated learning aggregation.
Semantic Classification
Content
0424 Confidential Computing — content pending enrichment.