TECH NEWS | VAST brings confidential AI to sensitive data

VAST Data has launched DataEnclave, a confidential AI runtime designed to let enterprises run AI models against sensitive data while protecting both the data and proprietary model weights during processing.

DataEnclave is part of the VAST AI Operating System and VAST DataEngine. Built with NVIDIA Confidential Computing, it uses hardware isolation and cryptographic attestation to verify the environment before model weights and sensitive data are decrypted.

The system addresses a security gap in AI computing: while encryption can protect data while stored and transmitted, AI models must eventually be decrypted in accelerator memory to perform inference.

“The last gap that’s actually existed until very recently is actually the memory, the high bandwidth memory that now sits on these accelerator cards, on these GPUs,” said John Mao, VAST vice president for global business development.

DataEnclave uses three main mechanisms: hardware-isolated execution, verify-before-decrypt attestation and independent key control.

Under the system, the model provider can retain control of the keys used to decrypt its model weights, while the enterprise retains control of its own data keys.

“If it is authorized to run, then there’s a key that’s basically exchanged to be able to decrypt the model and execute as you expect,” Mao said.

The approach is aimed at organizations that cannot move sensitive information to external AI services, including government agencies and other environments with strict data-security or sovereignty requirements.

“Obviously, there’s also air-gapped environments, classic air-gapped environments, like government agencies, that can never do that,” Mao said.

DataEnclave supports connected and fully air-gapped deployments. For a completely isolated environment, Mao said a hardened appliance would be required.

“In the completely air-gapped model, you would need a hardened appliance, effectively, that would have to be deployed on-prem,” he said.

VAST said DataEnclave can use attestation services based on the Cloud Native Computing Foundation’s Trustee project. It can also work with Fortanix’s confidential AI infrastructure for fully sovereign deployments.

The system records attestation events, key releases and enclave lifecycle actions in a tamper-proof audit trail. VAST said this allows organizations to determine what ran, where it ran and under which verified policy.

The secure runtime also extends to AI agents through VAST AgentEngine, providing isolated environments and controls over the data, systems and tools available to agents.

“Models are becoming a resource the operating system has to manage, the same way it manages data,” said Renen Hallak, founder and CEO of VAST Data.

The launch includes model builders such as Cohere, CrowdStrike, Deepgram, Factory, Fundamental, NVIDIA and TwelveLabs. VAST also listed Cisco, Supermicro and Lenovo as server partners, and NScale, Buzz and Sharon AI as AI cloud partners.

VAST acknowledged that confidential computing adds some performance overhead because of the additional processing required during inference. Mao said the impact depends on model size and that the company plans to publish benchmark results.

Full disclosure: All news articles published on the TechSabado website are written by human journalists, unless otherwise specified. Final text editing is also performed by human editors, with artificial intelligence (AI) used only to assist with additional grammar and style guide corrections..

————————————————————————-
TEN YEARS OF TECHSABADO!

PLEASE LIKE our FACEBOOK PAGE and SUBSCRIBE to OUR YOUTUBE & SPOTIFY channels.

WATCH TECHSABADO ON OUR YOUTUBE CHANNEL:

WE ARE ALSO ON SPOTIFY

WATCH OUR OTHER YOUTUBE CHANNELS:

PLEASE LIKE our FACEBOOK PAGE and SUBSCRIBE to OUR YOUTUBE CHANNEL.

WE ARE ALSO ON SPOTIFY

WATCH OUR OTHER YOUTUBE CHANNELS:

PLEASE LIKE our FACEBOOK PAGE and SUBSCRIBE to OUR YOUTUBE CHANNEL.

Jing Garcia