Explainability Drift in AI Models Explained in 60 Seconds | When Explanations Quietly Break

Explainability drift in AI models happens when the *reasons* behind a model’s predictions change over time, even if accuracy still looks fine. In this 1-minute glossary video, you’ll learn why explanations that once made sense can slowly stop matching what the model is actually doing under the hood. We’ll cover a plain-English definition, a simple mental model, a practical example, and why explainability drift matters for audits, regulation, and trustworthy AI. What you’ll learn: - What "explainability drift" means in modern AI systems - How a model’s decision logic can shift while metrics s…