The use of technology to help solve crimes is well established and advancing. But using technology to predict crimes is something different and a subject that will inevitably prove to be controversial, especially in in light of ‘profiling’ different sections of society.
Back in 2012, the New Orleans Police Department began a project with a data mining company called Palantir Technologies. The program was kept secret from the public, according to The Verge, and most of the City Council were oblivious to its use, Moreover, the use of the analytics was not revealed to the judiciary.
The basis of the partnership was for the police to use Palantir’s network-analysis software to identify potential aggressors as well as victims of violence. Palantir has patented a crime-forecasting system and also sells software to foreign intelligence services, which is used to predict the likelihood of individuals to commit terrorism.
Speaking with Laboratory Roots, Political Consultant James Carville is quoted as saying: “No one in New Orleans even knows about this, to my knowledge.” As the scheme became public in March 2018, this led to an adverse reaction from the public and the City Council, causing outgoing Mayor Mitch Landrieu to state that his office would not renew its pro bono contract with Palantir.
There is likely to be legal fall-out from the revelation, especially in questioning the validity of some convictions. The Verge reports further that this could well include filings from lawyers who represented people accused of membership in gangs.
Concerns with predictive policing technology include charges of racial profiling, where civil liberties groups state that the analytics for crunching data are often outdated, wrong or biased. A second is with privacy, such as putting people with no criminal convictions onto so-called ‘heat lists’ of potential criminals. Analysts also raise concerns about systems drawing too simplistic cause-and-effect conclusions, drawing inferences beyond the statistical capabilities of the software. This also links to concerns about an over-reliance upon such technology, ignoring the necessary part to be played by humans in drawing inferences from the data.
Other parts of the U.S. use different types of predictive policing technology and they have been relatively open about the use of the digital transformation of policing. An example is Pittsburgh’s “predictive policing” program and an equivalent scheme in New York City. Companies involved in such schemes include New York-based KeyStat Inc. and PredPol, a Santa Cruz, California-based company.
