Skip to content

v2.3.0 - Assurance & Accountability Profiles

Latest

Choose a tag to compare

@GLOBAL-AI-GOVERNANCE GLOBAL-AI-GOVERNANCE released this 30 Aug 05:30
· 8 commits to 524b787f435d2efabdb83d7c1d356398a50c47ce since this release

v2.3.0 — Assurance & Accountability Profiles

This release adds three optional profiles to the existing Decision Pack architecture: Auditable AI (1.0.0-rc.2), Quantum-AI Convergence (0.1.0, synthetic research only), and CISO AI Risk & Accountability (0.1.0, experimental).

All use a common deterministic Profile Result contract, a registry-backed gag profile command, explicit authority_effect: NONE, and the same accountable-human boundary. A profile may reference an existing Decision Pack and its digest but cannot rewrite it. The existing handoff can carry the profile-result path and digest in evidence_refs without changing its non-authorizing effect.

The browser adds a local-only renderer for existing profile results; it is not another evaluator. All examples are synthetic. This release does not provide certification, production authorization, universal safety, legal or regulatory determination, quantum advantage, real quantum navigation, operational Stateful Revocation, or autonomous approval.