Skip to content

Latest commit

 

History

History
53 lines (36 loc) · 2.58 KB

File metadata and controls

53 lines (36 loc) · 2.58 KB

ACR Framework Roadmap

This document outlines the development plan for the ACR Framework. For high-level governance and maintainer information, see the README.

Current Release: v1.0 (March 2026)

Status: Released

Delivered:

  • Core six-layer control plane architecture
  • Full specifications for all six pillars (Identity & Purpose Binding, Behavioral Policy Enforcement, Autonomy Drift Detection, Execution Observability, Self-Healing & Containment, Human Authority)
  • NIST AI RMF mapping and ISO/IEC 42001 alignment narrative
  • STRIKE threat model for AI-specific runtime threats
  • Implementation guide with three reference architectures (Kubernetes+OPA, AWS Serverless+Cedar, API Gateway+Custom)
  • Glossary, use cases, and FAQ
  • ACR telemetry schema specification (core event structure)

Planned Releases

v1.1 (Q2 2026)

Focus: Expanded implementation patterns, standardization, and compliance detail

  • Telemetry schema standardization — Formal versioning, optional fields, and migration guidance
  • Policy DSL requirements — Specification for a common policy authoring format
  • Drift detection requirements — Standard signals, baselines, and evaluation criteria
  • Deployment patterns — Additional patterns and decision guidance in a dedicated section
  • Compliance mappings:
    • SOC 2 Type II detailed mapping (control-by-control)
    • ISO/IEC 42001 detailed mapping (clause-by-clause)
  • Implementation guide — Additional runbooks and operational checklists

v1.2 (Q3 2026)

Focus: Multi-model orchestration and broader governance

  • Multi-model orchestration patterns — Governance across multiple agents and model endpoints
  • Federated governance — Cross-organization or cross-tenant ACR alignment
  • EU AI Act mapping — Alignment with EU AI Act requirements and risk tiers
  • Extended use cases — Deeper scenarios with sample policies and decision trees

v2.0 (2027)

Focus: Extensions for emerging AI architectures and regulatory modules

  • Emerging AI architectures — Adaptations for new agent patterns, tool-use norms, and infrastructure
  • Regulatory compliance modules — Optional, jurisdiction-specific compliance packs
  • Community and tooling — Mature community processes, link-check and doc CI, and optional documentation site

Contributing to the Roadmap

Suggestions for roadmap items are welcome. Open a GitHub Discussion with the tag roadmap or open an issue describing the proposed change and its rationale.