Technology
The ethical dilemmas of predictive policing in marginalized communities
Quick fact
Studies have shown that predictive policing algorithms often label poor or minority neighborhoods as high-risk even when controlling for actual crime rates, partly because they are fed historical arrest data that already reflects policing biases.
Why this is interesting
Imagine a software that tells police where to go to prevent crime – but it keeps sending them to the same poor neighborhoods. Why does that happen, and is it fair?