AI Policing: More Surveillance, Same Unaccountable Power
Mainstream media is buzzing about AI 'reshaping' policing, with tools from Axon and IBM generating reports and predicting 'crime patterns.' Axios claims this is about staffing shortages and public safety. What they omit is that police departments already disproportionately target Black and brown communities. Predictive policing isn't a neutral algorithm; it’s an automated feedback loop, amplifying
existing biases by training AI on data from historically over-policed neighborhoods. This isn't about saving man-hours; it's about institutionalizing systemic racism with algorithmic precision. So, while San Francisco police are piloting AI to draft reports (for 'citations and lower-level cases,' of course), one might wonder if these AI systems will be as 'efficient' at identifying patterns of
police misconduct or holding officers accountable. Or will the $6.6 billion AI law enforcement market merely serve to accelerate the carceral state's unchecked expansion, neatly packaging discrimination behind a veneer of technological neutrality? Apparently, 'bias deep within the criminal justice system' is an unfortunate side effect, not a core feature.