Linux Foundation sets 2026 confidential computing summit
Fri, 26th Jun 2026 (Today)
The Linux Foundation and OPAQUE have announced the schedule for Confidential Computing Summit 2026, which will focus on confidential computing and AI sovereignty.
The event in San Francisco will bring together enterprise executives, engineers and policymakers to discuss how confidential computing is being used to manage sensitive AI workloads across cloud and distributed systems. The programme includes keynotes and workshops on agent security, intellectual property protection, verifiable environments, infrastructure from cloud to edge, and data collaboration.
The summit comes as companies face growing pressure to process sensitive data in AI systems without breaching privacy, governance or security requirements. The Linux Foundation cited IDC research showing that 75% of organisations are adopting confidential computing, with 88% of respondents naming improved data integrity as the main driver.
Keynote speakers include executives and researchers from AMD, Google, Microsoft, the Technology Innovation Institute, UC Berkeley and OPAQUE. The wider agenda also features participants from Amazon, Meta, NVIDIA, Samsung Electronics, DigiCert and LY Corporation.
Confidential computing has come into sharper focus as businesses move from AI trials to broader deployment. The approach is designed to protect data while it is being processed, not just when it is stored or transferred. That has become more relevant as AI systems increasingly rely on regulated or proprietary information.
Jim Zemlin, Chief Executive Officer of the Linux Foundation, linked that shift to concerns over sovereignty and operational risk.
"As organizations move into full-scale AI deployment, there is a critical need to prioritize data sovereignty and risk mitigation," said Zemlin, Chief Executive Officer of the Linux Foundation.
He said the event reflects the work of the wider open-source and standards community on privacy and security in AI systems.
"Confidential Computing Summit brings together the community focused on building the open tools and standards necessary to ensure that privacy and security are not sacrificed for performance," Zemlin said.
Agenda themes
Several sessions will focus on what organisers describe as the trust layer for the agentic economy, referring to AI systems that can act more autonomously across business processes and datasets. That part of the programme includes a Microsoft presentation titled Trust Is the Next Bottleneck: Why the Agentic Economy Needs Confidential Computing, alongside talks from Samsung Research UK and Samsung Electronics, and a case study from LY Corporation on confidential virtual machines.
Another track will examine the operational side of scaling confidential infrastructure in Kubernetes and hypercluster environments. Sessions include WhatsApp Private Processing from Meta, a Google presentation on Kubernetes trusted execution environments for AI, and an NVIDIA talk asking whether Kubernetes is the right platform for confidential compute.
A further set of discussions will centre on attestation and hardware-rooted trust. Those sessions cover NVIDIA's work on attestation for AI hardware, DigiCert's view of the convergence between public key infrastructure and confidential computing, and an overview of the AWS Nitro System from Amazon.
OPAQUE, which is co-hosting the event, framed the issue less as a question of model performance than of operational trust. It argues that organisations need stronger proof that data, models and workflows remain protected while AI systems are running.
"AI adoption is no longer limited by model capability. It's limited by trust. As AI agents begin operating autonomously across sensitive systems and data, organizations need verifiable guarantees that their data, models, and workflows remain protected during execution, not just policies that assume compliance," said Aaron Fulkerson, Chief Executive Officer of OPAQUE.
That emphasis reflects a broader debate in the AI market, as businesses and regulators pay closer attention to how systems use sensitive information in live environments. Questions around runtime verification, hardware-backed isolation and governance controls are becoming more important as AI moves into sectors such as finance, healthcare and public services.
OPAQUE also linked confidential computing to the broader development of sovereign AI infrastructure.
"Confidential computing and Confidential AI are becoming foundational infrastructure for secure, sovereign AI at scale, and the Confidential Computing Summit brings together the leaders building that trust layer for the next generation of AI," Fulkerson said.
Sponsors for the summit include AMD, the Confidential Computing Consortium, Google, Microsoft, the Technology Innovation Institute and Tinfoil. The programme also highlights how much current activity in confidential computing is being shaped by a mix of cloud providers, chipmakers, software groups and academic researchers.
Among the keynote speakers, Ion Stoica stands out as both a Professor at UC Berkeley and a Co-Founder of Databricks, Anyscale and OPAQUE, underscoring the links between academic research and commercial work in this area. Other keynote speakers include Hugo Romero of AMD, Nelly Porter of Google, Mark Russinovich of Microsoft Azure, Dr. Najwa Aaraj of the Technology Innovation Institute and Fulkerson.
The schedule suggests the summit will focus less on broad claims about AI and more on the practical mechanics of securing data, systems and workloads as enterprises expand real-world use. IDC's finding that improved data integrity is the leading reason for adoption points to the same concern: companies want evidence that sensitive information remains protected when AI is in use.