NEW YORK – Super Protocol today announced that BE, a UK-based provider of AI infrastructure and services, has integrated Super Protocol’s confidential computing software to enable confidential LLM inference powered by NVIDIA Blackwell systems with NVIDIA Confidential Computing.
The integration combines BE’s UK compute infrastructure with Super Protocol’s complete software stack for confidential workload deployment, orchestration and verification. It is designed for enterprises with strict security, regulatory or data-governance requirements, protecting sensitive information during AI processing – including from the provider and privileged infrastructure administrators.
For these organisations, the alternatives have often been to restrict which information they use with external AI services or operate dedicated GPU infrastructure themselves. Confidential inference offers another route: advanced AI on provider infrastructure, with hardware-enforced protection and verifiable evidence of the software processing customer data.
Enterprise AI creates value by working with customer records, proprietary code, internal documents, research and sensitive operational data. Enterprises turn to inference providers for leading models, optimised performance and GPU capacity on demand. Sensitive workloads also require protection from the companies operating those services.
Consider a UK insurance company whose teams want to use AI to review claims, summarise supporting documents and check policy terms. These tasks involve customers’ identities, financial details, policy information, correspondence and other sensitive claims data – including medical information. The company needs access to advanced models without making that information accessible to the inference provider or its administrators.
Traditionally, encryption protects information in storage and during transfer. During ordinary AI processing, however, information becomes readable in memory, where privileged administrators or attackers who compromise the host may be able to access it.
Confidential Computing protects this processing stage inside hardware-isolated environments. Cryptographic verification provides evidence of the protected environment and approved software before sensitive information is released.
“Our enterprise customers need to protect sensitive data while running AI at scale, with fast responses, reliable availability and clear service commitments,” said Sath Ganesarajah, Founder and Chief Executive Officer, BE. “Super Protocol gives us the complete stack within our existing infrastructure, allowing us to deploy and scale protected inference without building a separate confidential computing platform. Customers can verify the service before sending sensitive information, giving them access to advanced AI with data privacy they can verify independently.”
NVIDIA Blackwell GPUs with Confidential Computing protect AI workloads and sensitive data during GPU processing. Paired with compatible confidential CPU environments, they create a protected execution environment that isolates processing from the host infrastructure and its privileged operators.
“As the UK expands its sovereign compute and inference capacity, enabling sensitive and regulated workloads is increasingly important,” said Anthony Hills, Regional Sales Leader for NVIDIA in the UK & Ireland. “NVIDIA Confidential Computing provides hardware-backed protection during AI processing. Super Protocol supplies the complete software stack that makes this protection deployable and verifiable across inference services, enabling providers such as BE to add verifiable confidentiality to UK-based sovereign AI infrastructure.”
Super Protocol turns CC-capable hardware into scalable confidential clusters. Its software layer verifies execution environments, deploys and orchestrates workloads, and protects communications as applications and capacity grow. BE provides the compute infrastructure and operates the confidential inference service within its existing service model.
Before submitting sensitive information, customers can verify that:
Requests are processed within the protected CPU/GPU environment, and results return securely. Application policies enforce access controls, logging and permitted outputs. BE maintains the service without accessing customer information inside the confidential environment.
For the insurer, this means using advanced models to help its teams review sensitive claims information on UK infrastructure, without owning dedicated GPUs or disclosing customer records to the inference provider. The same foundation can support finance, healthcare, defence and government, with sector-specific deployment and application policies.
“Confidential AI enables regulated industries to apply leading models to sensitive information while retaining control over their data,” said Nukri Basharuli, Founder and CEO of Super Protocol. “NVIDIA Blackwell delivers accelerated computing, while NVIDIA Confidential Computing, paired with compatible confidential CPU environments, protects data and workloads during processing. Super Protocol provides the complete software stack to deploy, verify and scale that protection within existing infrastructure. Our collaboration combines BE’s infrastructure and service expertise with Super’s confidential execution software, making advanced AI practical for sensitive workloads across finance, healthcare, defence and government.”
Sovereign AI strategies commonly focus on keeping data and processing within a chosen jurisdiction. Data residency addresses where information is processed, but not how it is protected while in use.
BE’s UK infrastructure, combined with Super Protocol, adds independently verifiable confidentiality to sovereign AI deployments. Customers can select infrastructure that meets their residency requirements and verify that their data is processed by approved AI software inside a protected environment.
Learn more about NVIDIA Confidential Computing at NVIDIA GTC Berlin on October 20-22, 2026, and explore case studies at superprotocol.com/case-studies.
BE is a UK-based provider of AI infrastructure and services with infrastructure projects in the UK, North America and other markets. Its services include GPU compute and AI inference, for enterprises with high-performance, security and data-governance requirements. BE also develops, owns and operates integrated power and compute infrastructure, bringing together energy supply, cooling and computing capacity to support large-scale AI workloads. Learn more at bec.uk.
Super Protocol provides a reusable Confidential AI software stack for creating and operating verified confidential environments at scale across an organisation’s chosen infrastructure. Super Swarm deploys and verifies applications in confidential environments across supported cloud, private and bare-metal systems, protecting sensitive data and workloads from infrastructure operators and privileged administrators during processing. Its Confidential AI layer makes these capabilities accessible to enterprises, AI developers and GPU cloud providers without requiring confidential computing expertise or building confidential environments from scratch.
