WIRED investigated a UK police predictive analytics system and found a messy deployment with some results that could not be trusted. The report comes as police forces experiment with AI and data tools for crime prediction and operational planning.
The stakes are high because flawed predictions can influence policing decisions, resource allocation, and public trust. When models are used in law enforcement, errors are not just technical issues; they can affect communities directly.
The story reinforces the need for transparency, validation, and limits around predictive systems in sensitive public-sector settings.