Confidential-Computing on CuraSechttps://curasec.metacog.co.kr/tags/confidential-computing/Recent content in Confidential-Computing on CuraSecHugoen-usMon, 24 Aug 2026 13:10:29 +0000CCAegis: Intra-Process Isolation via Arm CCA Granule Protection Tableshttps://curasec.metacog.co.kr/insights/2026-08-24-more-granular-less-trust-enforcing-intra-process-isolation-w/Mon, 24 Aug 2026 13:10:29 +0000https://curasec.metacog.co.kr/insights/2026-08-24-more-granular-less-trust-enforcing-intra-process-isolation-w/<ul> <li><strong>Engineer — Learn:</strong> Novel research on reducing TCB in confidential VMs by enforcing intra-process data isolation at the hardware level via Arm CCA — no running system changes needed today, but relevant for teams designing workloads on Arm-based confidential compute platforms.</li> <li><strong>SOC/IR — Skip</strong></li> <li><strong>Leader — Skip</strong></li> </ul>MOSAIC: Cryptographic Protocol for Confidential AI Inference Outsourcinghttps://curasec.metacog.co.kr/insights/2026-08-03-mosaic-masked-outsourcing-of-secure-ai-computations/Mon, 03 Aug 2026 15:12:30 +0000https://curasec.metacog.co.kr/insights/2026-08-03-mosaic-masked-outsourcing-of-secure-ai-computations/<ul> <li><strong>Engineer — Learn:</strong> Novel matrix-multiplication masking protocol enabling private transformer inference on untrusted servers, backed by LWE/LPN hardness assumptions; no action needed today, but worth tracking if evaluating secure enclaves or confidential computing architectures for AI workloads.</li> <li><strong>SOC/IR — Skip</strong></li> <li><strong>Leader — Skip</strong></li> </ul>Decentralized Confidential Computing for AI Workloads via Intel TDXhttps://curasec.metacog.co.kr/insights/2026-07-27-decentralized-compute-on-untrusted-hardware-using-intel-tdx/Mon, 27 Jul 2026 15:10:27 +0000https://curasec.metacog.co.kr/insights/2026-07-27-decentralized-compute-on-untrusted-hardware-using-intel-tdx/<ul> <li><strong>Engineer — Learn:</strong> Academic proposal combining Intel TDX, Intel Trust Authority, and NVIDIA Confidential Computing into a decentralized CVM platform — worth reviewing if you&rsquo;re evaluating confidential compute options for protecting model weights or training data, but no production tooling or immediate action follows from this paper.</li> <li><strong>SOC/IR — Skip</strong></li> <li><strong>Leader — Skip</strong></li> </ul>