PublishedGuidanceLAST REVIEWED · 30 AUG 2026
MC-1 Trust
The proposed continuous assurance layer for AI systems and autonomous agents.
What MC-1 Trust covers
MC-1 Trust is a development initiative for inventory, identity, authorization, containment, evidence, independent evaluation, and continuous verification. This guide describes the intended architecture; it does not assert production availability or existing AATS certifications.
Trust architecture
- MC-1 Core governs intelligence decisions.
- Agent Identity establishes who acts and who is accountable.
- Agent Guard and Guard Edge are intended to bound execution at authorized action points.
- The ColomboAI-led AATS draft is intended to be implementable without MC-1.
- The release-gated Trust service can store scoped evidence and recalculate status; automated source connections remain future work.
Current boundary
Was this page helpful?