ISO/IEC 22989
A common vocabulary for AI
ISO/IEC 22989 establishes AI terminology and describes concepts in the field of artificial intelligence. AI projects fail when people use the same AI words but mean different things.
The problem
Same words, different meanings.
Organisations talk about AI, machine learning, deep learning, models, training, inference, data, knowledge, prediction, classification, automation, agents, autonomy, human oversight, risk, bias, and trustworthiness.
Without a common language, business teams, developers, vendors, management, auditors, and regulators misunderstand each other. ISO/IEC 22989 helps create a shared AI vocabulary.
What I research
Concepts I make explicit in every project.
- AI terminology
- AI system concepts
- Machine learning concepts
- AI lifecycle understanding
- Human role in AI systems
- AI-assisted automation
- AI-generated outputs
- AI agent responsibilities
- AI risk language
- AI governance language
- AI documentation
- Stakeholder communication
Outcome
The organisation knows what AI is doing — and what it is not.
Standard vocabulary allows me to explain AI-enabled systems in a way that business, technical, and governance teams can understand. It makes clear what AI is doing, what it is not doing, where humans remain responsible, what data it uses, what outputs it produces, what risks must be controlled, and how it is monitored and improved.