Our methodology
The principles and approach underpinning our AI governance assessments.
Assessment philosophy
AIGC Australia takes a principles-based, evidence-driven approach to AI governance assessment. We evaluate not just whether governance documentation exists, but whether it reflects genuine organisational commitment and practical implementation.
Our methodology is designed to provide structured governance assessment to boards, procurement teams, and external stakeholders while supporting organisations in strengthening their governance maturity over time.
Assessment outcomes reflect governance maturity at the time of assessment and do not constitute regulatory approval.
Principles-based
Assessment against outcomes rather than prescriptive checklists
Evidence-driven
Decisions based on documented practices and implementation evidence
Maturity-focused
Recognition of governance as a continuous improvement journey
Our methodology
AIGC conducts governance maturity assessment using a five pillar model: accountability and ownership, risk identification and register integration, policy and control alignment, oversight and review mechanisms, and independent assessment. Each pillar is evaluated against a structured maturity scale to ensure consistent and defensible scoring.
Scoring is evidence based, supported by an evidence register, structured interviews, and documented independent peer review prior to maturity determination.
Pillar 1: Accountability and ownership
Assessment of governance accountability structures, role clarity, and ownership allocation for AI systems and decisions.
Pillar 2: Risk identification and register integration
Evaluation of AI risk identification processes and integration into enterprise risk management frameworks.
Pillar 3: Policy and control alignment
Assessment of policy frameworks and control mechanisms governing AI development, deployment, and operation.
Pillar 4: Oversight and review mechanisms
Evaluation of governance oversight processes, review cadence, and monitoring arrangements for AI systems.
Pillar 5: Independent assessment
Assessment of independent review mechanisms, assurance activities, and external validation processes.
Independence safeguards
Structural independence
Our assessment function operates independently from any advisory or consulting activities. Assessors are prohibited from providing implementation advice to organisations they assess.
- Separation between advisory and assessment functions. Where advisory services are provided, structural separation and cooling off safeguards apply prior to governance assessment
- No commercial relationships beyond the defined assessment engagement
- Assessor rotation to prevent familiarity bias
Conflict of interest management
All assessors must declare any relationships or interests that could compromise objectivity. Assessments are reassigned where potential conflicts are identified.
Commercial independence
Assessment outcomes are determined solely on evidence and criteria. Fee structures are fixed and do not depend on assessment outcomes.
Evidence-based decision making
Documentation review
We examine governance documentation including policies, procedures, risk assessments, and accountability frameworks. Documentation provides evidence of governance design and intent. Evidence sufficiency is assessed against defined criteria.
We recognise that documentation quality varies, and we assess substance over form.
Implementation evidence
Beyond documentation, we seek evidence that governance practices are implemented in practice. This includes review of decision records, audit trails, training records, and incident response histories.
Implementation evidence demonstrates that governance is operationalised, not merely documented, where sufficient evidence is available.
Stakeholder engagement
Where appropriate, we engage with key personnel responsible for AI governance to understand decision-making processes, accountability structures, and organisational culture. Interviews are conducted to supplement documentary evidence and do not replace documented controls.
Conversations provide context that enriches document-based evidence.
Consistency and quality assurance
Standardised framework
All assessments follow the same structured framework, ensuring consistent evaluation regardless of organisation size, sector, or assessor assignment.
Assessor calibration
Assessors participate in regular calibration exercises to ensure consistent interpretation and application of assessment criteria across all engagements.
Peer review
Assessment findings are subject to documented independent peer review prior to assessment outcome, providing an additional quality control layer.
Continuous improvement
We review methodology periodically to ensure responsiveness to evolving governance expectations, based on assessment experience and stakeholder feedback.
Framework alignment
Our methodology is informed by Australia's AI Ethics Principles and relevant international governance frameworks. We update our approach as Australian government guidance evolves.
Listing of these resources does not imply endorsement.
What assessments do not include
To maintain independence and clarity of purpose, our assessments specifically exclude:
- Implementation advice: We do not advise organisations on how to implement governance practices
- Technical testing: We do not conduct technical audits, penetration testing, or algorithmic testing of AI systems
- Legal compliance assessment: We do not assess legal compliance; assessment addresses governance practices
- Ongoing monitoring: Assessment reflects a point-in-time evaluation and does not constitute continuous oversight
Learn more about our framework
Explore the detailed assessment criteria and governance framework.