ATLAS

Attestation and Transparency Layer for AI Systems

Open Standard — v0.3

ATLAS specifies how AI systems produce cryptographic evidence that governance controls executed correctly. It defines governance domains, agentic AI controls, human-in-the-loop attestation, and conformance levels for healthcare, regulated deployments, and autonomous agents. The standard is vendor-neutral, encoding-agnostic, freely implementable, and designed for adoption through Protocol Profiles. ATLAS is maintained by GLACIS Technologies.

Invitation for Public Review and Comment
Crosswalks: NIST AI RMF· ISO 42001· EU AI Act· IETF RATS