In predicting policing ethics, what is necessary to ensure acceptability in practice?

• A. Unlimited data and no external review.
• B. Exclusive use for minor offenses.
• C. Validation and oversight to address privacy, bias, transparency, and accountability.
• D. Automatic deployment without evaluation.

Answer: (C)

Validation and oversight are essential to make predictive policing tools acceptable in practice. Validation checks that the model performs fairly and safely across different contexts and populations, while oversight provides governance to protect privacy, detect and mitigate bias, ensure results are explainable, and establish clear accountability for how predictions are used. Together they build trust, meet legal and policy requirements, and allow ongoing improvement. The other approaches fall short: unlimited data with no external review risks privacy breaches and biased outcomes; restricting use to minor offenses ignores broader ethical concerns and fairness across the system; deploying without evaluation can propagate harms without remedy.