When the police caught the Golden State Killer in 2018, it was remarkable for two reasons. First, many of the crimes committed took place decades earlier. Second, it’s one of the most compelling uses of AI in law enforcement.
Avoiding the gritty details of the case – AI’s use in crime detection, both preventative and reactive is now very much in vogue.
A year ago, Argentina announced the creation of an AI Security Unit, which is designed to use algorithms to analyse historical crime data and predict potential future crimes.
Meanwhile,the UK government launched the Concentrations of Crime Data Challenge to develop AI-based tools for predicting and preventing crime.
The private sector’s getting in on the action. A UK police force testing an AI tool from engineering firm Akoddis has reported that it can perform work equivalent to 81 years of detective analysis in 30 hours.
Startup Flock Safety reckons round the clock AI surveillance powered by camera bearing drones and cameras on every street would end crime in America. They may have a point. Though killjoys in civil liberties groups have raised a few questions.
Perhaps the best known company in the space, Palantir shows the risks and rewards. The company cashes huge cheques by partnering with government agencies, including police departments. But a few years back, LAPD ended its use of the companies predictive policing programs amid public outcry, with evidence mounting it reinforced harmful policing patterns.
Balance is needed, but there’s no doubt we’ll see more AI in policing and crime prevention. Policing is increasingly expensive for governments, and private companies would love to get their hands on some of the funding. Data is hard to come by, but a study from 2000 estimated global public policing at approximately $194 billion. Other estimates scratch at $500bn. Either estimate is enough to lure the attention of startups looking for a market during the AI gold rush.
