AiGovOps Foundation · Topic
What is OVERT?
OVERT means Observable Verification Evidence for Runtime Trust. It is an open standard for producing independently verifiable runtime evidence that declared governance policies, security controls, and oversight actions actually executed in AI systems, including AI agents’ tool calls, without exporting protected content. The Foundation is co-steward of OVERT with CHAI.
Key facts
- OVERT means Observable Verification Evidence for Runtime Trust.
- OVERT defines independently verifiable runtime evidence for governance policies, security controls, and oversight actions.
- Its scope includes AI agents’ tool calls without exporting protected content.
- OVERT 1.1 was published 11 June 2026 and has a royalty-free covenant.
- The Foundation is co-steward of OVERT with CHAI; the agreement is signed.
- Glacis Technologies, Inc. is the maintainer.
Frequently asked questions
What does OVERT stand for?
OVERT stands for Observable Verification Evidence for Runtime Trust.
What does the OVERT standard do?
It defines independently verifiable runtime evidence that declared governance policies, security controls, and oversight actions actually executed in AI systems.
Does OVERT cover AI agents’ tool calls?
Yes. The standard’s scope includes AI agents’ tool calls without exporting protected content.
Who co-stewards OVERT?
The AiGovOps Foundation is co-steward of OVERT with CHAI, the Coalition for Health AI.
What is the current OVERT version?
OVERT 1.1 was published 11 June 2026 and has a royalty-free covenant.
Updated 2026-10-09 · Ken Johnston & Bob Rapp
Take the next step
Continue with one practical next step from the AiGovOps Foundation.
Try Beacon, an OVERT-conformant runtime →