What happens if all distances in KNN are almost equal?

KNN loses discriminative power because no meaningful neighborhood structure exists.

When distances converge (common in high-dimensional spaces), KNN cannot distinguish between near and far points. This results in unstable predictions and effectively random neighbor selection.