en.Wedoany.com Reported - In May 2026, Dell Technologies and Intel jointly addressed the "Securing the AI Factory" strategic dialogue, systematically outlining a full-stack security architecture for AI infrastructure. They identified hardware root of trust, zero trust governance, and post-quantum cryptography as the three core pillars of AI factory security. Mike Ferron-Jones, Head of Platform Security and Integrity Marketing at Intel, engaged in a public discussion with several Dell executives on security architecture against the backdrop of agentic AI migrating to on-premises deployment. Both parties disclosed their technical roadmap, spanning from confidential computing to quantum-safe cryptographic acceleration.
The conclusions of Intel's "2026 Platform Security Report" served as the technical starting point for the discussion: security must be built into the chip itself; otherwise, any software controls layered on top are untrustworthy. Ferron-Jones pointed out directly that the CPU is the foundation of the entire security stack, and choosing a CPU is the first security decision an enterprise makes. Based on this, Intel categorizes data center security capabilities into four major areas—protecting the platform from boot-level attacks, protecting data through confidential computing, enforcing security via hardware, and accelerating encryption without performance trade-offs. Dell has integrated this architecture into its PowerEdge server line, utilizing Intel Xeon processors to provide confidential computing and root-of-trust boot verification, complemented by PowerScale secure storage protocols and the PowerProtect data protection suite to safeguard AI models and data at rest.
Confidential computing was defined as the hardware-based implementation of zero trust architecture within AI infrastructure. Through Intel SGX and TDX technologies, Intel creates hardware-enforced Trusted Execution Environments, providing organizations with cryptographic verification that attests to the location and operational state of sensitive workloads. This capability targets sectors with strong compliance requirements, including finance, healthcare, and government. Ferron-Jones explicitly stated that data is only released into a confidential TEE via customer-controlled encryption keys, and whether facing compliance or traditional cybersecurity threats, the data keys always remain under customer control. Intel is also collaborating with NVIDIA to develop reference architectures that extend confidential AI environments to GPU-accelerated workloads.
Post-quantum cryptography represents a deep area of collaboration between the two parties. Dell Global CTO John Roese had previously warned publicly that quantum computing will break all current encryption systems, and agentic AI, with its higher-value data, exacerbates the risk. Ferron-Jones noted that the most threatening attack today is not a real-time quantum attack, but "harvest now, decrypt later"—where attackers collect encrypted data to decrypt it once quantum capabilities mature. Intel is initiating the transition to quantum-safe encryption with Xeon 6 processors, planning for platform-wide cryptographic quantum safety by 2029. Ferron-Jones emphasized that data at rest can be encrypted today using quantum-safe AES-256 algorithms, and the instruction acceleration built into current Xeon CPUs makes this transition possible without significant performance overhead. Concurrently, Dell announced in March 2026 that its commercial PCs would come pre-equipped with quantum-safe firmware protection.
Dell and Intel position the security challenges of AI factories as an infrastructure proposition, not a matter of add-on tools. Dave Vellante, Chief Analyst at theCUBE Research, pointed out that AI factories cannot rely on traditional security methods and require a control plane capable of governing data, models, and agents in real time. Krista Case added that global cybersecurity data breaches increased by 40% year-over-year in 2026, IoT malware attacks surged by 124%, and the data layer has become the primary attack target in AI environments. Nearly three-quarters of organizations are already at a stage of AI maturity or full integration, while approximately one-quarter cite security exposure as their top risk.
Dell's Data Orchestration Engine, a low-code data automation tool released in March 2026, automates data discovery, preparation, and governance, transforming scattered data into AI-ready, high-quality datasets. Dell is also continuously expanding its AI Factory portfolio through Generative AI Solutions in collaboration with Intel, where Intel Gaudi3 AI accelerators paired with Dell high-performance servers provide scalability for generative AI workflows. Dell's U.S. headquarters is located in Round Rock, Texas. The full-stack "secure by design" approach formed by Dell and Intel provides continuous control plane capabilities for the complete lifecycle of an AI factory, from the underlying hardware to the upper-layer data pipelines.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
