LAPD audit reveals dangers of high-tech policing

An audit of the Los Angeles police department is raising questions about new technologies law enforcement isusing nationwide with little oversight.

Last week, the department’s internal auditors, prompted by a community backlash, published online a review of the LAPD’s data-driven policing strategies and recommended more transparency, consistency and oversight of the programs. Los Angeles has been a leader in using new technologies such as artificial intelligence, social networks and big data to aid police work.

However, those new approaches have been controversial, and criticized for violating civil rights and discriminating against minority groups. Experts have also questioned whether they deliver enough value given their financial costs, which can exceed hundreds of thousands of dollars.

The audit examined programs that targeted locations and people for additional policing. A chronic-offender program tracked individuals to see if they were committing violations. A point system was implemented to determine who should be on the list. Points were given if a person was a gang member or on parole, had a history of violent crime, was arrested with a handgun, or had “quality police contact” — a term that wasn’t defined in department documents.

However, the audit found nearly one in five people on the list had not received any points.

Last year, theLAPDsuspended the chronic offenders list because of civil-rights concerns and inconsistencies in how it was administered. The department continuesto usethe other new policing technology. The audit also cautioned against drawing strong conclusions on whether any of the new approaches lessened crime.

Two other LAPDprograms analyzed data to increase police presence in areas crime seems likely to take place – a technique called predictive policing. One program uses gun-related crime data, which it feeds into the department’s Palantir platform. Palantir offers a centralizedlocation to collect and analyze police data, including license-plate scans, rap sheets, foreclosed properties and traffic tickets. Another of the predictive-policing programs, which was from the software company Predpol, relies on 10 years of crime data to suggest locations.

Related article This is the first police officer charged with a federal hate crime in at least 10 years

Los Angeles police chief Michel Moore told the city’s police commission last week that he believed data-driven policing would improve community safety, and that his department supported the audit’s recommendation.

“I’m thankful that this report has come,” Moore said.

Experts say that Los Angeles’ public review shows the value of audits, which hold law enforcement accountable but are a rarity even as departments increasingly adopt new technologies.

“If they embrace this idea of accountability and transparency, they can actually showcase the good, deal with the bad and move things forward in a really constructive way,” said Andrew Ferguson, a law professor at the University of District of Columbia and author of “The Rise of Big Data Policing.”

It’s estimated that more than 50 police departments use technologies similar to theLAPD. A 2014 study found that 38% of police departments said they used predictive policing, and 70% planned to within the next two to five years.

Los Angeles is believed to be the first department to have audited its use of these cutting edge technologies.

Auditors say their value to the departmentsis offering an independent and objective look.

“We don’t sweep things under rug. We just say it how it is,” said Kristine Adams-Wannberg, president of the Association of Local Government Auditors, which was not involved in the Los Angeles audit. “If a police bureau has been touting their record to say hey, we’re doing a really good job. An audit can come in to verify that.”

Adams-Wannberg said that Los Angeles has become a leader in police audits, dating to the fallout of a 2001 settlement with the Department of Justice over civil rights violations.

Chicago, one of the largest police departments nationwide that’s using technologies similar to Los Angeles, is trying an alternative to audits. It’s partnered with the University of Chicago researchers to get feedback on its programs.

“All police departments should seek mechanisms of accountability and transparency. Accountability through academic studies and reports is beneficial,” said Ferguson, the law professor. “But there is no real substitute for local democratic accountability in a public and transparent forum.”

“I don’t know that you can have a stronger validation or efficacy model than an independent Ivy League university,” said Anthony Guglielmi, a spokesman for the Chicago police department. “They literally sit in the rooms with us.”

In early 2017, Chicago’s police department set up small offices in two troubled neighborhoods for police officers and university data analysts to work together to better target police resourcesin order toreverse a rising trend of gun violence. The offices rely on a suite of technologies, including automatic gunshot detection, which alerts officers of shootings before 911 calls. They also use data analytics and a software program, HunchLab, that attempts to predict when and where crimes are likely totake place.

A recent university analysis found a significant reduction in shootings in one of the neighborhoods when compared to other districts that didn’t have one of the data-focused offices.

The results are promising, but no one has studied which of the new tools drove the crime reduction. HunchLab, the type of predictive policing program that’s been controversial, isn’t being examined on its ownby the university.

Even when there are promising results from the new systems,critics argue that more audits and transparency around the programs is critical to make sure the technology is actually effective and doesn’t raise civil rights issues.

“Police tend to give technological solutions a benefit of the doubt,” said Eric L. Piza, a professor at the John Jay College of Criminal Justice?who has studied predictive policing, and believes more needs to be done to make sure the technology works well enough to justify the financial costs. “It’s so so rare that police take a longer look.”