Palantir, the data-surveillance software employed by the Los Angeles Police Department (LAPD), was found to use demographic data such as tattoos, scars, license plates and physical appearance as a way to target criminals, according to training documents obtained by Buzzfeed News.
The software, which LAPD has used for more than 10 years as a part of the Los Angeles Strategic Extraction and Restoration (LASER) initiative, was originally designed to be used in military war zones.
Described as “evidence-based policing on steroids,” Palantir creates a large database of people in the city suspected of criminal activity, making it easier for law enforcement to search for information on individuals.
In the LAPD training documents shown due to Buzzfeed’s Freedom of Information Act Request, Palantir Gotham categorizes people by their gender and race but also their scars, tattoos, gang membership, friends and family or license plate number. This means that people who are innocent can also be captured–and kept on the system forever.
Once a police officer successfully searches a person in the database, the officer can see their address and contact information as well as any existing warrants or mugshots. That information can be used to label a person as someone with the propensity to commit a crime in the future.
Every day from 2011 to 2019, LAPD officers used Palantir to identify 12 “chronic offenders”: those considered having the most propensity to commit a crime.
A person was chosen as a chronic offender on a point system: a violent crime in the past two years, being on parole, gang membership or arrest due to possession of a handgun were worth five points each. Quality police contracts were worth one point each. In 2017, they also had to list five to ten additional “backup” people to the list of 12.
To make matters worse, they could only identify these “chronic offenders” in something called a LASER zone, which comprises 40 areas that LAPD declared to be at high risk for criminal activity.
Although LAPD claims that they no longer create chronic offender bulletins, such prolonged use of them has led to negative feedback from the community.
An LAPD inspector general report showed that although Los Angelos has a population that’s 52 percent white, 49 percent Latino and 9 percent Black, chronic offenders were 53 percent Latino and 31 percent Black, representing disproportionality in those chosen as chronic offenders.
“That means that many folks who end up in the Palantir system are predominantly poor people of color, and who have already been identified by the gaze of police,” said Andrew Ferguson in the Buzzfeed article.
The use of Palantir in order to identify chronic offenders as a method of data-driven policing “often results in officers using the veneer of objectivity to back up their hunches,” said Sarah Brayne, a sociologist who’s been studying Palantir at LAPD for two years.
According to the inspector general report in the article, LAPD was inconsistent in their choice of chronic offenders as “44 percent were never arrested for a violent or gun-related crime,” and “10 percent of Chronic Offenders had zero contacts with police — meaning zero arrests and zero mentions in-field interviews.”
According to Buzzfeed, it’s unknown if people ever knew they were listed as a chronic offender. Even if they were alerted, typically by mail or in-person by an officer, the only way to remove their information from Palantir would be through a court order.
While LAPD reported a 23 percent drop in crime in 2016, the inspector general report showed that the violent crime rate remained the same or even got worse in 6 out of 13 LASER zones.
Although LAPD ended the LASER program in April 2019 and recently launched the Data-Informed, Community Focused Policing Program, it continues to use Palantir software and is strikingly similar to LASER.
The only difference is that it actually widens the scope of police power. In the article, LAPD said that LASER’s surveillance was for those arrested for violent or gun-related crimes. Under the new program, those merely suspected of nonviolent or property crimes can be surveilled as well.
“The federal government shouldn’t be spending money on unproven surveillance software or crime prediction programs that target Black and Hispanic Americans and don’t actually reduce crime,” said Senator Ron Wyden (D-OR) in the article.
The new program leaves people suspecting that police tactics “are more of the same, even if they are packaged as new and scientific,” said Jamie Garcia, a member of the advocacy group “Stop LAPD Spying Coalition.”
This summary was prepared by TCR news reporting intern Emily Riley.