Concept Drift
A change in the statistical relationship between input features and the target variable over time, causing a previously accurate model to become less reliable and requiring retraining or intervention.
In plain language
When the real world changes in ways that make your AI's rules outdated. What counts as spam, fraud or a good customer today is different from a year ago. Your AI was trained on old patterns and doesn't work as well anymore.
Why this matters
Concept drift is critical to AI governance because models trained on historical data may not perform as expected when the underlying business environment shifts. Your organisation should establish monitoring processes to detect performance degradation and define decision rules for model retraining or replacement.
Relevance
ImplementationIdentifying and responding to concept drift requires ongoing monitoring and retraining discipline. Without this, your deployed AI systems will gradually lose effectiveness and may introduce new risks.
Related terms
Putting concept drift into practice in your organisation?
Ready to transform your AI strategy?
Partner with Australia's AI strategy and governance specialists. From adoption roadmaps to ISO 42001 audit readiness.