OPAQUE, AMD, Intel and others back TRACE, an open standard for hardware-enforced AI runtime verification

OPAQUE, a company focused on confidential AI, announced TRACE (Trust, Runtime Attestation, and Compliance Evidence), describing it as an open, portable standard intended to help verify what AI systems are doing while they run. The effort is aimed at “AI agents and other confidential workloads,” with the stated goal of making it easier to demonstrate trust and compliance during execution rather than relying only on pre-deployment settings or audits.

TRACE is pitched as a hardware-enforced governance record. In the announcement circulated through Yahoo Finance, OPAQUE said TRACE is designed to capture and carry compliance evidence that is generated at runtime, supported by runtime attestation. Runtime attestation is the process of proving, during execution, that a system is operating under the expected security state, rather than only claiming it once at install time.

The initiative includes participation from major semiconductor and security ecosystems. Alongside OPAQUE, the announcement named AMD, Intel, and TII, described as industry leaders collaborating on TRACE. The involvement of chip makers matters because hardware-backed attestation and isolation features are often the basis for proving an execution environment’s integrity.

For Intel specifically, the disclosure in the market report frames the company as one of the collaborators behind an effort to advance runtime verification for AI workloads. The practical implication is that the standard is being positioned to work across different deployments by making the evidence portable, so enterprises and developers could potentially verify runtime behavior consistently across environments.

From a broader technology standpoint, TRACE reflects a growing push in confidential computing and AI security toward stronger assurance mechanisms. As AI agents move from static assistants to systems that take actions in real time, vendors and customers have increasingly asked for assurances that the agent is running as intended, under the right controls, and can provide evidence that those controls were active during the time of execution.

Still, the announcement does not spell out many specifics in the market report, including technical implementation details, which attestation mechanisms TRACE would map to, or how compliance evidence would be structured for different regulatory or contractual requirements. It also does not provide deployment timelines or confirm which platforms and hardware features are immediately supported.

What OPAQUE and its named partners did disclose, however, is the direction: TRACE is intended to be open and portable, and the company is framing it as a way to “advance runtime verification” for AI agents and confidential workloads. For now, the exact scope of “compliance evidence,” how it is verified by third parties, and whether it is ready for production use remain unclear based on the information in the published report.

The next items to watch are concrete deliverables that typically follow these kinds of standards announcements: reference implementations, compatibility statements for specific hardware and software stacks, and guidance on how attestation and the compliance record are expected to work in real deployments. Investors and customers are also likely to look for whether additional industry members join and whether major cloud and enterprise platforms support the standard.