Data Residency Enforcement in Multi-Region AI Deployments
Infrastructure must enforce where data lives, not just policy documents.
Lorena Sundqvist
Editor at Large
Lorena Sundqvist
Lorena Sundqvist covers zero-trust and provable data privacy, confidential computing on gpus and confidential ai inference at scale for Private AI Review.
10 stories
CPU-GPU Confidential Channel Establishment with AMD SEV-SNP
Researchers detail how CPUs and GPUs authenticate each other under AMD's confidential computing.
Lorena Sundqvist
Latency and Throughput Overhead of TEE-Protected Inference
SGX's memory ceiling makes confidential VMs the safer choice for large models.
Lorena Sundqvist
Model Weight Protection Inside Confidential Inference Servers
Hardware enclaves protect model weights where policy and access controls cannot.
Lorena Sundqvist
Verifiable Transparency Logs for Model Provenance
Cryptographic logs make it impossible to hide who trained a model or when it changed.
Lorena Sundqvist
Attestation Token Replay and Freshness Attacks
Attackers replay valid tokens to bypass security checks that depend on current device state.
Lorena Sundqvist
Post-Quantum Cryptography Migration for AI Inference APIs
Quantum computers could decrypt your inference traffic today if captured now.
Lorena Sundqvist
Confidential Containers with Kata Containers and TEE Backends
VM-based TEEs offer practical confidentiality for container workloads without code changes.
Lorena Sundqvist
Side-Channel Attack Surfaces in SGX Enclaves
OS-level attackers can reconstruct enclave secrets from a single execution trace.
Lorena Sundqvist
NVIDIA H100 Confidential Computing Mode Architecture
Hardware encryption protects AI model weights on GPUs from privileged attackers.
Lorena Sundqvist