Submitted to JMIRO — March 2026
The CAIOS Standard
Clinical AI Oversight Specification — an evidentiary framework submitted for peer review, for responsible AI governance in healthcare.
What is CAIOS
An evidentiary framework, not a regulator.
CAIOS proposes what should be provable when AI assists a clinical decision: the artefacts, attestations, and audit trails a practice can maintain to demonstrate considered governance to insurers, boards, and medico-legal reviewers. It is currently under review at JMIRO; the platform does not depend on its publication.
Buyer-facing assessment leans on the existing references the audience already trusts — RANZCR Chapter 9, TGA guidance for Software-as-a-Medical-Device, AHPRA codes of conduct, ACSQHC quality standards, and DISR responsible-AI principles. CAIOS organises the evidence; it does not replace the source.
Five domains
The evidentiary framework
DOMAIN 01
Clinical Governance
Organisational structures, accountability frameworks, and decision-making authority for AI oversight. Defines who is responsible, how decisions are escalated, and what governance bodies must exist.
AccountabilityEscalationOversight Bodies
DOMAIN 02
Risk Management
Identification, assessment, and mitigation of risks arising from clinical AI deployment.
Risk matrix
Likelihood
DOMAIN 03
Regulatory Compliance
Alignment with TGA, RANZCR, and emerging international AI governance frameworks.
TGARANZCR Ch.9NHMRCISO 14971
DOMAIN 04
Evidence & Audit
Immutable documentation of governance decisions, attestations, and evidentiary artefacts.
evidence-vault/
risk-assessment.pdf
governance-policy.pdf
audit-attestation.pdf
audit-pack-v2.zip
DOMAIN 05
Continuous Monitoring
Ongoing surveillance of AI system performance, drift, and compliance posture.
Compliance score94%
JanFebMar
Score your practice against CAIOS
Professional governance gap analysis delivered as a board-ready PDF.
Get your assessmentDevelopment
How CAIOS was built
2023
Research begins
Systematic review of AI governance gaps in clinical radiology across Australia and New Zealand.
2024
Framework drafted
Five-domain evidentiary framework developed from clinical observation, regulatory analysis, and medicolegal precedent.
2025
Expert refinement
Framework refined through expert consultation, regulatory analysis, and alignment with TGA guidance and RANZCR standards.
Mar 2026
Submitted to JMIRO
CAIOS submitted for peer review to the Journal of Medical Imaging and Radiation Oncology.
When a reviewer asks how you governed your AI, ‘we did our best’ is not evidence.”
Review-ready evidence
Structured artefacts prepared for insurer, board, regulatory, and medico-legal review.
Insurer readiness
Demonstrate governance posture to medical indemnity insurers. Structured evidence of AI oversight is what insurers and underwriters ask for — CAIOS maps directly to those requirements.
Regulatory alignment
Pre-aligned with TGA, RANZCR Ch.9, and NHMRC frameworks. When regulation catches up, you're positioned for alignment.
See where your practice stands.
Free readiness assessmentThe difference
With CAIOS vs without.
Without CAIOS
- Scattered Word docs and shared drives
- No audit trail for AI-assisted decisions
- No proof of governance when it matters
- Vulnerable during formal external review
- Insurers see unquantified risk
With CAIOS
- Sealed decision packs per clinical event
- Hash-verified chain of governance evidence
- Timestamped artefacts linked to every AI tool
- Court-grade proof of responsible oversight
- Insurer-ready governance posture on demand
Built for alignment with
Pursue Governance Verification.
Advisor-reviewed verification that your governance evidence is mapped to the CAIOS framework. Available as an add-on for Scale and Enterprise.