#1 Data - Instead of collecting tools like license plate readers and gunshot detection systems from multiple companies, police departments are shifting towards a monopoly of popular sensors and cameras using AI to analyze data. Along with the surveillance state, potential data breaches and theft threaten our privacy.
#2 Police Reports - Regulation, consistency, biases from pre-existing crime data, and producing false information are some of the consequences of AI-created police reports, too. These reports are critical in trials that can take place months or even years after an event.
#3 Virtual Reality - Virtual reality training systems allow police to practice interactions, but can't capture soft skills and nuances in human behavior. Hacking raises privacy concerns. Cybersickness, which functions like motion sickness, could lead to interactions with disgruntled cops struggling to accomplish tasks properly. And there's more.
These tools have an environmental cost. In 2019, researchers estimated that training one AI model generates 626,000 pounds of carbon dioxide, over five times the amount a car produces in its lifetime. As police budgets balloon, what will the financial cost be?
When an MIT Technology Review reporter attended the International Association of Chiefs of Police conference, he noticed a theme: "If your department is slow to adopt AI, fix that now. The future of policing will rely on it." But before we believe the hype, let's question whether this future will keep us safe.