Linux Foundation Backs New TRACE AI Security Standard 

A new open standard that’s designed to provide evidence of AI agent activity is being contributed to the Linux Foundation, where it will be developed under a vendor-neutral governance model. 

The standard, called TRACE (Trust, Runtime Attestation and Compliance Evidence), was collaboratively developed by OPAQUE (which contributed it), AMD, Intel, Microsoft, and the Technology Innovation Institute (TII). In a Tuesday press release, the Linux Foundation said that TRACE is tackling a very relevant challenge with today’s prevalence of AI usage: proving that sensitive data in organizations is being handled “according to policy” by AI agents and open weight models.

“TRACE creates a standardized, hardware-enforced governance record that binds together the runtime environment, software, policies, data classifications and tool usage into a portable, cryptographically verifiable artifact that travels with the workload across clouds and confidential computing environments,” according to the Linux Foundation. 

Why is TRACE Needed?

TRACE defines a Trust Record that contains evidence about an agent’s execution, including the model that ran, what environment it was running in, what classification of data was involved, and what type of governance rules were supposed to be applied to the model's execution. 

In its press release, OPAQUE (a cybersecurity company focused on confidential computing and privacy-preserving technologies), one of the organizations behind TRACE, said that the standard aims to address a governance challenge for organizations, which was best encapsulated by the OpenAI-Hugging Face intrusion: “documented policies and sandbox configurations do not, by themselves, prove which controls remained in force or what a system actually did during execution.”

That governance gap is not trivial. According to a January 2026 report from Deloitte, three-quarters of companies plan to deploy Agentic AI within two years; but only 21 percent of those companies reported having a mature model for agent governance.

What is TRACE?

Rather than making another security framework from scratch, TRACE brings together six existing industry standards into a “common evidence layer,” or a way of documenting what an AI system does. Those standards (including Secure Production Identity Framework for Everyone, or SPIFFE; Supply-chain Levels for Software Artifacts or SLSA; and Remote Attestation Procedures, or RATS) already address areas like software supply chain, remote attestation procedures, workload identity, and hardware-based trust. 

“The initial specification focuses on today's AI agent architectures while providing a foundation that can evolve with future multi-agent and confidential computing deployments,” according to the Linux Foundation. 

The project has seen early interest: the TRACE reference library recorded nearly 135,000 PyPI downloads within 10 weeks of its initial introduction at the Confidential Computing Summit in June. 

Linux Foundation and the Impact

Under the Linux Foundation, TRACE will receive vendor-neutral governance for managing the project. The technical workstream will be hosted by the Coalition for Secure AI (CoSAI).

“The widespread adoption of autonomous systems requires independent, cross-platform proof of operational integrity,” said Jim Zemlin, CEO of the Linux Foundation. "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."