Research

Open research on trustworthy AI for live infrastructure. Protocol design, red team findings, and authenticated evidence under an explicit collector trust model: architectural enforcement rather than model-policy compliance.

Current research artifacts live in the specification, the security page, and the Architecture and Trust Model.

Archive: The Trust Problem (March 2026) — preserved as a historical document; superseded by the pages above.

Get Involved

VIRP is open. The Internet-Draft is public. The implementation is Apache 2.0.

Contact Us