An AI-powered facial recognition system is deployed in a cit…
An AI-powered facial recognition system is deployed in a city. Due to technical limitations in the training data—specifically, poor lighting conditions and lower resolution images for people with darker skin tones—the system exhibits significantly lower accuracy (higher false positive/negative rates) for this demographic. Since this unfair discrimination stems from a technical flaw in the data or hardware/software tooling, what category of bias is this?