Glossary
AI governance glossary
Clear definitions of common terms used in AI governance and Australian privacy law.
A
- AI governance
- The policies, roles, processes and controls an organisation uses to make sure its use of AI is safe, lawful, ethical and aligned with its objectives. Our AI governance services
- AI management system (AIMS)
- A structured set of policies and processes for managing AI across an organisation. ISO/IEC 42001 sets out requirements for one. ISO/IEC 42001 explained
- AI policy
- An organisation’s rules for using AI. It usually covers approved tools, data handling, human oversight, accountability and incident reporting. AI policy essentials
- AI register (AI inventory)
- A living list of the AI tools, models and automated decision systems an organisation uses, with an owner, purpose, data and risk rating for each. How to build an AI register
- Algorithmic bias
- Systematic errors in an AI system’s outputs that unfairly favour or disadvantage particular people or groups, often inherited from training data or design choices.
- APP 1.7
- A new part of Australian Privacy Principle 1, commencing 10 December 2026, requiring privacy policies to describe certain decisions made or substantially supported by computer programs using personal information. APP 1.7 explained
- Australian Privacy Principles (APPs)
- The 13 principles in the Privacy Act 1988 that govern how covered organisations and agencies collect, use, disclose, store and secure personal information.
- Automated decision-making (ADM)
- Using computer programs, such as rules engines, scoring models or AI, to make decisions, or to do things substantially and directly related to making them. Check if a decision is in scope
E
- Explainability
- The ability to describe, in terms people can understand, why an AI system produced a particular output or decision.
G
- Generative AI
- AI that creates new content, such as text, images, audio or code, in response to prompts. Large language models are a common example.
H
- Hallucination
- When a generative AI system produces information that sounds plausible but is false or unsupported.
- Human-in-the-loop
- A design where a person reviews, approves or can override an AI system’s output before it takes effect. This does not automatically take a decision outside the privacy transparency rules.
I
- ISO/IEC 42001
- An international standard specifying requirements for establishing, maintaining and continually improving an AI management system. It is not generally mandatory in Australia. Read more
L
- Large language model (LLM)
- A type of AI model trained on large amounts of text to understand and generate language. Most AI chat assistants are built on one.
M
- Model drift
- A decline in an AI model’s performance over time as the data or conditions it sees in practice change from what it was trained on.
O
- OAIC
- The Office of the Australian Information Commissioner. It is the national regulator for privacy and freedom of information, which publishes guidance on AI and privacy.
P
- Personal information
- Information or an opinion about an identified individual, or an individual who is reasonably identifiable, whether true or not, and whether recorded or not.
- Privacy impact assessment (PIA)
- A systematic review of how a project or system affects individuals’ privacy, with recommendations to manage or reduce the risks.
R
- Responsible AI
- Designing, buying and using AI in ways that are safe, fair, transparent, accountable and respectful of people’s rights.
S
- Sensitive information
- A category of personal information given extra protection under the Privacy Act, including health, genetic and biometric information, racial or ethnic origin, and religious beliefs.
- Shadow AI
- Use of AI tools by staff without the organisation’s knowledge or approval. Managing shadow AI
V
- Vendor (third-party) AI risk
- Risks that arise from AI products and services supplied by other companies, including data use, security, accuracy and accountability. Vendor due diligence questions
Definitions are simplified for general understanding and are not legal definitions. Missing a term? Let us know.