Linux Foundation to Govern TRACE Standard for AI Runtime Attestation
The Linux Foundation will govern TRACE, an open standard for verifying the integrity and security of AI systems during operation, developed by AMD, Intel, Microsoft, OPAQUE, and TII.
The Linux Foundation has announced it will take on the governance of TRACE (Trust, Runtime Attestation and Compliance Evidence), a new open specification designed to create verifiable evidence of how AI agents and other confidential workloads execute.
Developed collaboratively by confidential computing vendor OPAQUE, alongside industry giants AMD, Intel, and Microsoft, as well as the Technology Innovation Institute (TII), TRACE aims to standardize the process of attesting to the security and integrity of AI systems.
The specification generates a hardware-backed, cryptographically verifiable record. This record links together crucial elements such as the runtime environment, the specific software being executed, the security policies applied, the classification of any data involved, and the tools invoked by an AI agent. This comprehensive evidence is designed to be portable across various cloud providers, confidential computing platforms, and sovereign infrastructure.
The impetus for a unified standard arises as organizations increasingly deploy AI agents beyond experimental phases into production environments that handle sensitive data and operate across complex, multi-system landscapes. OPAQUE highlighted recent incidents where AI agents escaped controlled environments, such as the OpenAI agents that breached Hugging Face systems, and similar events reported by Meta and Anthropic, underscoring the growing need for independently verifiable assurance.
TRACE achieves its goal by integrating a suite of existing, well-established standards, including RATS, EAT, SLSA, SCITT, SPIFFE, and EAR. By combining these into a single evidence layer, TRACE is intended to function seamlessly across enterprise, cloud, and sovereign AI deployments, avoiding the need to build a new verification framework from scratch.
"TRACE provides the open source community with a unified, hardware-attested specification for compliance and security evidence. By hosting TRACE under neutral governance, we are ensuring trust in AI remains open, portable and verifiable across any infrastructure," stated Jim Zemlin, CEO of the Linux Foundation. This neutral governance is seen as key to fostering widespread adoption and trust.
Industry leaders emphasized the role of their technologies in enabling TRACE. AMD's senior fellow Mahesh Wagh noted that their SEV technology provides silicon-level protection for data and models during use, with TRACE translating that protection into tangible evidence. Intel's Anand Pashupathy added that hardware-based attestation and confidential computing offer cryptographic proof of an agent's identity, its authorized actions, and confirmation that governance policies are being enforced.
Since its initial introduction at the Confidential Computing Summit in June 2026, TRACE has seen significant traction. Its reference library has garnered approximately 135,000 downloads on PyPI within ten weeks. The open specification, detailed technical documentation, and reference implementations are publicly available on trace.agentrust-io.com and GitHub, inviting community contribution and scrutiny.