01 What happened
Cohere announced the public beta of Confidential Computing on Model Vault, running on NVIDIA H100 GPUs in Microsoft Azure. The system uses hardware isolation and remote attestation to let customers verify the inference environment before sending sensitive data.
02 Key details
- Cohere Model Vault Confidential Computing is now in public beta on Microsoft Azure with NVIDIA H100 GPUs.
- The system protects GPU memory, uses OHTTP to encrypt request payloads, and performs remote attestation before data release.
- Customers control attestation policies and can verify measurements independently.
- Cohere plans to open-source the stack and add NVIDIA Blackwell B200 support before general availability.
03 Why it matters
Enterprises need verifiable assurance that AI inference remains confidential from cloud and platform operators. Cohere's approach makes the trust boundary explicit and measurable, reducing reliance on implicit infrastructure trust.
04 Who it matters to
AI security teams, compliance officers and developers deploying confidential workloads in Azure.
Original source