Australian Government Mandatory AI Guardrails · vVoluntary AI Safety Standar…

Australian Government Mandatory AI Guardrails

Australian Government Mandatory AI Guardrails — 90% covered.

10 requirements · 9 enforced · 0 designed · 0 advisory · 0 deferred.

Source: Department of Industry, Science and Resources (DISR), 'Safe and responsible AI in Australia' Proposals Paper for introducing mandatory guardrails for AI in high-risk settings + the Voluntary AI Safety Standard (September 2024). 10 guardrails.

By category

CategoryReqsEnforcedDesignedAdvisoryDeferredCoverage
Accountability & governance 1 1 0 0 0 100%
Risk management 1 1 0 0 0 100%
Data governance & security 1 1 0 0 0 100%
Testing & monitoring 1 1 0 0 0 100%
Human oversight 1 1 0 0 0 100%
Transparency 1 1 0 0 0 100%
Contestability 1 1 0 0 0 100%
Supply-chain transparency 1 1 0 0 0 100%
Record-keeping 1 1 0 0 0 100%
Stakeholder engagement 1 0 0 0 0 0%

Every requirement → the KYE™ artefact that enforces it

IDTitleStatusKYE™ enforcement
au-ai-guardrails.G1 Guardrail 1 — Establish, implement and publish an accountability process including governance, internal capability and a strategy for regulatory compliance enforced audit_events: kye.compliance.attestation.v1, kye.purpose.request.v1
engines: internal, internal
constitution_refs: constitution/12-PURPOSE-PERMISSION.md, constitution/00-INDEX.md
au-ai-guardrails.G2 Guardrail 2 — Establish and implement a risk management process to identify and mitigate risks enforced audit_events: kye.assurance.risk_assessment.v1, kye.purpose.admissibility.v1
engines: internal, internal
constitution_refs: constitution/13-RESILIENCE-LOOP.md
au-ai-guardrails.G3 Guardrail 3 — Protect AI systems and implement data governance measures to manage data quality and provenance enforced audit_events: kye.evidence.tool_call.v1, kye.evidence.decision_map.v1
engines: internal, internal
constitution_refs: constitution/31-DATA-GOVERNANCE-PACK.md
au-ai-guardrails.G4 Guardrail 4 — Test AI models and systems to evaluate performance and monitor the system once deployed enforced audit_events: kye.resilience.drift.detected.v1, kye.resilience.measurement_record.v1
engines: internal, internal
constitution_refs: constitution/13-RESILIENCE-LOOP.md
au-ai-guardrails.G5 Guardrail 5 — Enable human control or intervention in an AI system to achieve meaningful human oversight enforced audit_events: kye.purpose.admissibility.v1, kye.evidence.decision_map.v1
engines: internal
constitution_refs: constitution/36-GOVERNEDUI.md, constitution/12-PURPOSE-PERMISSION.md
au-ai-guardrails.G6 Guardrail 6 — Inform end-users regarding AI-enabled decisions, interactions with AI and AI-generated content enforced audit_events: kye.evidence.decision_map.v1, kye.evidence.pack.v1
engines: internal, internal
constitution_refs: constitution/13-RESILIENCE-LOOP.md
au-ai-guardrails.G7 Guardrail 7 — Establish processes for people impacted by AI systems to challenge use or outcomes enforced audit_events: kye.replay.proof.v1, kye.evidence.pack.v1
engines: internal, internal
constitution_refs: constitution/13-RESILIENCE-LOOP.md, constitution/21-DELEGATED-AUDITABILITY.md
au-ai-guardrails.G8 Guardrail 8 — Be transparent with other organisations across the AI supply chain about data, models and systems to help them effectively address risks enforced audit_events: kye.risk.authority_register.v1, kye.evidence.tool_call.v1
engines: internal, internal
constitution_refs: constitution/21-DELEGATED-AUDITABILITY.md
au-ai-guardrails.G9 Guardrail 9 — Keep and maintain records to allow third parties to assess compliance with the guardrails enforced audit_events: kye.evidence.pack.v1, kye.replay.context_seal.v1
engines: internal, internal
constitution_refs: constitution/30-AUDIT-WORM-RETENTION.md
au-ai-guardrails.G10 Guardrail 10 — Engage stakeholders and evaluate their needs and circumstances, with a focus on safety, diversity, inclusion and fairness out-of-scope audit_events: kye.compliance.attestation.v1
constitution_refs: constitution/49-UNIVERSAL-ENGAGEMENT-RAIL.md