An AI model for predictive policing is challenged by civil liberties advocates because it uses historical arrest data to predict future crime. The primary AI ethics concern with this application is:
Select an answer to reveal the explanation.
Short Explanation
Here's the deal: you're dealing with a feedback loop. Historical arrest data reflects over-policing, not necessarily real crime, so sending more police to those communities creates more arrests and reinforces the bias.
Full Explanation
The correct concern is that historical arrest data encodes systemic biases in policing, and using it to predict future crime creates a feedback loop that reinforces and amplifies those biases by directing more police to already over-policed communities. Historically over-policed communities have higher recorded crime, not necessarily higher actual crime. Predicting from that data sends more police there, producing more arrests that feed back as confirmation into training data, amplifying racial and socioeconomic disparities rather than reflecting objective crime prediction. The data-minimization concern is secondary. The explainability concern is important but not the primary ethics concern. There is no universal 85% accuracy threshold for public safety AI applications.