In May and June 2013, when the New Orleans murder rate was the sixth highest in the United States, the Orleans district prosecutor delivered two racketeering indictments against dozens of men accused of belonging to two violent gangs. of drug traffickers from Central City, 3NG and the 110ers. Members of both gangs were accused of committing 25 murders and several attempted assassinations and armed robberies.
Subsequent investigations by the Bureau of Alcohol, Tobacco, Firearms and Explosives, the Federal Bureau of Investigation and local agencies produced new RICO allegations, including that of a 22-year-old man named Evans "Easy" Lewis, a member of a gang called 39ers who were accused of participating in a drug distribution network and several murders.
According to Ronal Serpas, the head of the department at that time, one of the tools used by the New Orleans Police Department to identify gang members such as 3NG and the 39ers came from the Silicon Valley company, Palantir. The company provided software to a secret NOPD program that traced people's ties to other gang members, described criminal histories, analyzed social networks and predicted the likelihood of people committing violence or becoming victims. As part of the discovery process at Lewis' trial, the government handed over 60,000 pages of documents detailing evidence compiled against him from confidential informants, ballistics and other sources, but did not mention NOPD's association with Palantir, according to a source. familiar with the 39ers trial.
The program began in 2012 as a partnership between the New Orleans Police and Palantir Technologies, a data mining company founded with seed money from the CIA venture capital firm. According to interviews and documents obtained by The Verge, the initiative was essentially a predictive surveillance program, similar to the "priority list" in Chicago that aims to predict which people are likely to be drivers or victims of violence.
The association has been extended three times, and the third extension will expire on February 21, 2018. The city of New Orleans and Palantir have not answered questions about the current status of the program.
Predictive surveillance technology has proven to be highly controversial wherever implemented, but in New Orleans, the program escaped public notice, in part because Palantir established it as a philanthropic relationship with the city through the NOLA For Life program. Thanks to its philanthropic status, as well as to the "strong" New Orleans model of government, the agreement never went through a public procurement process.
"Until now, nobody in New Orleans knows anything about this."
In fact, the key members of the city council and the lawyers contacted by The Verge had no idea that the city had any kind of relationship with Palantir, nor did they know that Palantir used his program in New Orleans to market his services to another agency of law enforcement for a multi-million dollar contract.
Even James Carville, the instrumental political agent in achieving Palantir's collaboration with NOPD, said the program was not public knowledge. "As far as I know, no one in New Orleans knows anything about this," Carville said.
More than half a decade after the partnership with New Orleans began, Palantir has patented at least one crime forecasting system and sold similar software to foreign intelligence services to predict the likelihood of people committing terrorism.
Even within the law enforcement community, there are concerns about the possible implications of civil liberties of the kind of individualized prediction developed by Palantir in New Orleans, and whether it is appropriate for the American criminal justice system.
"It is not the right tool for the application of local and state law."
"They are creating a list of targets, but we are not going to look for Al Qaeda in Syria," said a former police official who observed Palantir's work first-hand, as well as the company's sales arguments for predictive surveillance. The former official spoke on condition of anonymity to freely discuss his concerns with data mining and predictive surveillance. "Palantir is a great example of a ridiculous amount of money spent on a technological tool that may have some application," said the former official. "However, it is not the right tool for the application of local and state law."
Six years ago, one of the most secretive and powerful technology firms in the world developed a controversial intelligence product in a city that has served as a neoliberal laboratory for everything from charter schools to radical housing reforms since Hurricane Katrina. Because the program was never public, important questions about its basic operation, risk of bias and decorum in general were never answered.
Founded in 2004 by Alexander Karp and Peter Thiel (the company's largest individual shareholder), Palantir Technologies' rapid rise to the fifth most valuable company in the world has been fueled by lucrative contracts with Pentagon and State intelligence services. United, as well as security services. In recent years, Palantir has sought to expand its merger and data analysis business to the private sector, with mixed success.
Prediction is not a new territory for Palantir. Since at least 2009, Palantir was used by the Pentagon to predict the location of improvised explosive devices in Afghanistan and Iraq, a wartime risk assessment program that does not address concerns about the civil liberties that accompany surveillance Predictive individualized. Its commercial software platform, Metropolis, also uses, according to reports, predictive analysis to help companies develop consumer markets and streamline investments. But before 2012 with the New Orleans program, there is no record publicly available that Palantir has ventured into predictive surveillance.
Interest and investment in predictive surveillance technology accelerated after 2009, when the National Institute of Justice began issuing grants for pilot projects in anticipation of crimes. Those grants underpin some of the best known and most controlled predictive policing efforts in Chicago and Los Angeles. Programs vary, and algorithms are often proprietary, but all aim to ingest large amounts of data (geography, criminal history, weather, social network histories) and make predictions about individuals or places that might be involved in a crime. In the following years, many start-up companies have struggled to monetize the crime fighting method, most notably PredPol, a California startup whose award contracts have failed after an initial bombing of advertising in the early 1990s. 2010
Before 2012 with the New Orleans program, there are no records publicly available that Palantir has ventured into predictive surveillance
As more departments and companies began experimenting with predictive surveillance, government-funded research cast doubt on its effectiveness, and independent academics discovered that it may have an unequal impact on poor communities of color. A 2016 study designed the reverse engineering PredPol algorithm and found that it replicated the "systemic bias" against communities of color over police and that historical crime data did not accurately predict future criminal activity. One of the researchers, a Michigan State doctorate candidate named William Isaac, had not previously heard of the New Orleans association with Palantir, but recognized the data mapping model at the heart of the program. "I think that with the data they are using, there are serious doubts about its predictive power, we have seen very little about its ability to predict violent crimes," said Isaac.
According to interviews and documents obtained by The Verge, Palantir first approached New Orleans in 2012 through a well-known intermediary: James Carville, the agent of the power of the Democratic Party and architect of Bill Clinton's successful presidential campaign in 1992. Carville is a paid advisor to Palantir, whose relationship with the data mining company dates back at least to 2011.
"I am the sole promoter of that project."
In an interview, Carville told The Verge that he was the impetus for the collaboration between Palantir and New Orleans. "I am the sole promoter of that project, I was totally my idea," Carville said, adding that he and Palantir's executive director, Alex Karp, flew to New Orleans to meet with Mayor Landrieu. "For me, it was a case of morality, young people were fighting each other, and the public was not as involved as they should have been."
The documents describing Palantir's relationship with New Orleans describe the company's role as "pro bono" and philanthropic. In 2015, Palantir mentioned his work in New Orleans in his annual philanthropy report, which characterizes the effort as a collaborative "network analysis" for law enforcement and other interested parties in the city.
Carville's statements about a public radio station in the Bay Area four years ago clarify how Palantir's relationship with the city occurred. In a January 2014 appearance on the KQED Forum talk show, Carville and his wife Mary Matalin praised Palantir's work in New Orleans as a major driver of the two-year reduction in the homicide rate.
"The CEO of a company called Palantir, the CEO, a guy named Alex Karp, said they wanted to do a charity job, and what did I think, I said, we have a really horrible crime rate in New Orleans," Carville told the presenter from KQED Forum, Michael Krasny, without mentioning his professional relationship with Palantir. "Then he went down and met with our mayor … both had the same reaction to the absolute immorality of the young people who killed other young people and society did nothing about it, and we could, at no cost to the city, begin to integrate data and predict and intervene as to where these conflicts were going to come in. We have probably seen a third of a reduction in our homicide rate since this project began. "
Matalin, who is also a political consultant, made it clear to Krasny that the prediction work being done by Palo Alto with NOPD was both a prototype and could potentially sweep innocent people.
"Unless you're the cousin of a drug dealer who went wrong, you're going to be fine."
"We are a kind of prototype," said Matalin. "Unless you're the cousin of a drug dealer who went wrong, you're going to be fine."
Ronal Serpas, chief of police in New Orleans from 2010 to August 2014, recalled his initial contact with Palantir staff during a meeting initiated by Mayor Landrieu's office. "They came and discussed the kind of work they do in theaters of war, the kind of work they do in other parts of the world," Serpas said during an interview at his office at Loyola University. "My impression was that Palantir was also interested in trying to develop products that could predict the crime."
The relationship between New Orleans and Palantir was finalized on February 23, 2012, when Mayor Landrieu signed an agreement granting New Orleans free access to the company's public sector data integration platform. Licenses and technical support for the Palantir law enforcement platform can reach millions of dollars annually, according to an audit by the Los Angeles County Sheriff's Department.
In January 2013, New Orleans would also allow Palantir to use its law enforcement account for the LexisNexis Accurint product, which is comprised of millions of searchable public records, court records, licenses, addresses, telephone numbers and contact information. social networks. The company also obtained free access to criminal and non-criminal data of the city in order to train its software for crime prevention. Neither the residents of New Orleans nor the key members of the city council whose job it is to monitor the use of municipal data knew Palantir's access to a large amount of data.
Palantir has a history of secrecy, and New Orleans is not the only instance of the company that conducts business with government agencies through nonprofit partner organizations, bypassing the public procurement process.
Palantir provides data analysis and integration for the Los Angeles Police Department, but the agreement was made through the Los Angeles Police Foundation rather than the Los Angeles Police Department itself. In New York, the company's contract was not disclosed by the city's comptroller for security reasons (the Police Department does this with surveillance equipment contracts) and was never taken to the municipal council for approval. Palantir's work with the New York police only became public when documents leaked about his tumultuous relationship with the country's largest police force to BuzzFeed journalist William Alden.
In New Orleans, according to extensive reports from The Verge, the office of Mayor Landrieu, the city attorney and the NOPD appear to be the only entities that know the work of the company in the city. The key members of the city council were unaware of Palantir's work in New Orleans until The Verge approached them.
The Palantir association would probably have received more scrutiny from the city council if it had been detailed in a budget, but council approval is not necessary for such a program. The structure of city government in New Orleans is based on a "strong mayor" model in which the council has no approval authority over contracts or policies for the city's police department.
Cities across the country have recently begun to deal with the question of whether municipalities should regulate data sharing and privacy. Some cities such as Seattle and Oakland have passed laws that establish committees to develop guidelines and conduct oversight, while others such as New York are discussing what role councils should play in relation to privacy in the digital age.
"I do not think there's anyone in the council who would say they knew this had happened."
Several civil and criminal lawyers who are heavily involved with the New Orleans criminal justice system were also unaware of any predictive surveillance efforts by the NOPD. Multiple criminal lawyers had never seen any of Palantir's analytical products as part of any discovery material given to them in the course of the test cases, although such an analysis would typically be required to be given to a defense attorney if it had been used as part of the investigation. an investigation of NOPD.
Jason Williams, president of the city council of New Orleans and a former defense attorney, reviewed the documentation of Palantir's collaboration with NOPD at the request of The Verge. Williams said he had never heard of the company's involvement with NOPD.
"I do not think there is anyone on the board who would say that they knew this had happened because this was not part of any of our budget allocations or our oversight," Williams said in an interview during a council meeting.
Williams, who also served as a criminal court judge before his election to the city council in 2014, said he did not necessarily oppose the use of data-based methods to help New Orleansans at risk.
"My main concern would be how it was used in my city, if it were used to identify marginalized people who are at risk of being harmed, to avoid damage, I will have a completely different appreciation of what I will have if this system is used badly. "
"It's almost as if New Orleans were hiring their own version of the NSA."
Councilwoman Susan Guidry, who chairs the council's criminal justice committee and has been in office since 2010, was also unaware of the New Orleans association with Palantir and the crime prevention work of NOPD. When he was shown the documentation of the NOPD program, Guidry told The Verge that he had never seen it before.
The Verge shared program documentation with a group of New Orleans civil rights lawyers. None previously knew about NOPD's prediction work, although one had heard rumors that Palantir was collaborating with NOPD, and they were concerned about the secrecy surrounding the program.
"It is especially disturbing that this level of intrusive investigation into the lives of ordinary residents is practically kept secret," said Jim Craig, director of the Louisiana office of the Roderick and Solange MacArthur Justice Center. Craig, who reviewed the program's documentation at the request of The Verge, compared the effort of predictive surveillance with signal intelligence work. "It's almost as if New Orleans was hiring its own version of the NSA to conduct a 24/7 surveillance of the lives of its people," Craig said. The authorities, he believes, have kept the program secret because it would provoke widespread outrage. "Right now, people are outraged with the traffic cameras and have no idea that this data mining project is underway," Craig said. "The South is still a place where people value their privacy very much."
Nicholas Corsaro and Robin Engel are two professors at the University of Cincinnati who conducted a recent evaluation of the New Orleans violence reduction strategy for which Palantir was used, and helped design a NOPD gang database from which bases the forecast model of Palantir. Both Engel and Corsaro ignored the predictive police efforts of New Orleans, their involvement in Palantir, or even the fact that the database they designed was fueling the program. "Trying to predict who will do what based on last year's data is pure crap," Corsaro said in an interview.
Palantir sometimes referred publicly to his work in New Orleans. However, none of Palantir's public presentations about the program that The Verge could identify went into detail about individualized crime forecasting, scraping social media data or using social media analysis for crime prediction. . Instead, the company represented its work in New Orleans as "developing a better understanding of the propensity to violent crime and designing specific interventions to protect the most vulnerable populations in the city."
"Trying to predict who will do what based on last year's data is bullshit."
In a public presentation praising the effectiveness of his work in New Orleans, Courtney Bowman, a civil liberties engineer from Palantir strongly involved with the company's work with NOPD, acknowledged that excessive secrecy could deepen the breach between police forces and the police communities. During a May 6, 2016 presentation at the DataEdge conference at the UC Berkeley School of Information, Bowman said: "These kinds of programs only work if the community is comfortable with the degree of application of this type of information and if They know how information is used. "
The City of New Orleans and Palantir rejected requests for comments on how their association was formed, and what kind of input had other elected officials and the public on the data mining company's predictive surveillance efforts.
Ronal Serpas, who led NOPD when the partnership with Palantir began, said he believed the city council and the general public should have been informed of the decision of the police department to participate in predictive surveillance with Palantir. The role of local legislatures and governing bodies in overseeing the exchange of government information is far from resolved, but Serpas believes that agreements with firms like Palantir deserve greater scrutiny.
"For me, it's something that certainly requires a view, it requires a look," Serpas said.
Although neither the staff of Palantir nor the current New Orleans officials would talk about the day-to-day running of the crime forecasting initiative, the documents obtained by The Verge, the external studies and the recollections of the old main Serpas offer a portrait of how the Predictive surveillance The beta test has worked in the last six years.
The model of prediction of Palantir in New Orleans used an intelligence technique called social network analysis (SNA) to establish connections between people, places, cars, weapons, addresses, social network publications and other evidence in previously isolated databases. Think of the analysis as a practical version of a painting by Mark Lombardi that highlights the connections between people, places and events. After entering a query term, such as a partial license plate, nickname, address, telephone number or social media handles or publishes, the NOPD analyst will review the information tracked by the Palantir software and determine which persons are most at risk of committing violence or becoming a victim, depending on their connection to known victims or offenders.
The data on individuals come from information extracted from social networks and NOPD criminal databases for ballistics, gangs, probation and parole information, jail phone calls, service calls, the central case management system (it is say, every case that NOPD had registered) and the deposit of letters of field interviews of the department. The latest database represents every documented encounter that NOPD has with citizens, even those that do not result in arrests. In 2010, The Times-Picayune revealed that Chief Serpas had ordered that the collection of field interview cards be used as a measure of the performance of officers and districts, resulting in more than 70,000 field interview cards completed in 2011 and 2012 The practice resembled the NYPD "stop and frisk" program and was instituted for the express purpose of gathering as much information as possible about the inhabitants of New Orleans, regardless of whether or not they committed a crime.
A partial license plate, a nickname, an address, a phone number, a social media identifier
Then, NOPD used the list of possible victims and perpetrators of violence generated by Palantir to target individuals for the CeaseFire program in the city. CeaseFire is a form of the decades-old carrot and stick strategy developed by David Kennedy, a professor at John Jay College in New York. In the program, the police inform potential criminals with criminal records that they know of their past actions and will prosecute them as much as possible if they re-offend. If the subjects choose to cooperate, they are "called" to a required meeting as part of their conditions of probation and probation and are offered job training, education, possible job placement, and health services. In New Orleans, the CeaseFire program runs under the umbrella of NOLA For Life, which is Mayor Landrieu's favorite project that he has funded through millions of dollars from private donors.
According to Serpas, the person who initially ran the analysis of social networks in New Orleans from 2013 to 2015 was Jeff Asher, a former intelligence agent who joined the NOPD of the CIA. If someone had been shot, Serpas explained, Asher would use Palantir's software to find people associated with them through field interviews or social media data. "This data analysis shows the names and connections between people in FIs [field interview cards] at traffic stops, at victims of reports, informing victims of crimes together, whatever the case may be. Information is valuable to anyone who is doing research, "said Serpas.
According to Palantir's own documentation, Asher and his colleagues conducted social network analysis of each victim of a fatal or non-fatal shooting in New Orleans from 2011 to 2013. Through this technique, which Asher dubbed "The NOLA model," the City devised a list of approximately 3,900 people who were most at risk of being involved in gun violence because of their connection to a fencer or previous victim. "We can identify 30-40% of the shooting victims," Asher said at the 2014 internal conference of Palantir. Asher rejected repeated requests for an interview.
Theoretically, Asher's approach is substantially influenced by the research of Andrew Papachristos, a Yale professor who tracked the violence as if it were a communicable disease that spreads through association networks. However, given that his work was cited as the academic underpinning of crime forecasting models employed by PredPol and the Chicago Police Department, Papachristos has sought to distance his research from those methods.
Once NOPD generated its list of likely shooters and victims, the police department and social service providers – for the "carrot" side of NOLA For Life – would select people who were incarcerated or under judicial supervision for a "meeting" of calls ".
Mayor Landrieu's office promotes the program frequently, referring to him as an essential part of New Orleans' criminal justice policy. Palantir also claimed credit: "We are helping to break the cycle of violence" in New Orleans, read a passage in the 2015 report of the Philanthropy Engineering company. But its real impact is not clear.
Of the 308 people who participated in calls from October 2012 through March 2017, seven completed vocational training, nine completed "paid work experience", none completed a high school diploma or a GED course, and 32 were employed at some point through referrals Fifty participants were arrested after their call and since then two have died.
On the contrary, the application of the law vigorously pursued its end of the program. Since November 2012, when the new Multiple Agencies Gang Unit was founded, until March 2014, the accusations of organized crime increased: 83 alleged gang members in eight gangs were charged in the 16-month period, according to a presentation internal of Palantir.
The city devised a list of approximately 3,900 people who were most at risk of being involved in armed violence
Incoming calls dropped precipitously after the first few years. According to city records, eight call groups went into effect from 2012 to 2014, but only three took place in the following three years. Robert Goodman, a New Orleans native who became a community activist after completing a prison sentence for murder, worked as a "responder" for the city's CeaseFire program until August 2016, discouraging people from participating in violence of retaliation. Over time, Goodman noted a greater emphasis on the "stick" component of the program and greater control over non-punitive aspects of the program by the city council that, in his view, undermined the intervention work. "It's supposed to be run by people like us instead of the city trying to dictate how this should look," he said. "While they are not putting resources in the hoods, nothing is going to change." You're just putting band-aids.
After the first two years of Palantir's participation with NOPD, the city suffered a sharp drop in assassinations and armed violence, but it was short-lived. Even the former head of NOPD, Serpas, believes that the preventive effect of calling dozens of people at risk – and accusing dozens of them – began to diminish.
"When we finish with almost nine or ten accusations with about 100 defendants for federal or state RICO violations of killing people in the community, I think we caught the attention of many people in that criminal environment," Serpas said, referring to the accusations of organized crime. "But over time, it must have disappeared because before I left on August 14, we could see that things were starting to fall apart."
Nick Corsaro, profesor de la Universidad de Cincinnati que ayudó a construir la base de datos de pandillas de NOPD, también trabajó en una evaluación de la estrategia CeaseFire de Nueva Orleans. Descubrió que la disminución general de los homicidios en Nueva Orleans coincidió con la implementación del programa CeaseFire en la ciudad, pero los vecindarios de Central City a los que se dirige el programa "no tuvieron disminuciones estadísticamente significativas que correspondían con la fecha de inicio de noviembre de 2012".
Dicho llanamente, el estudio no confirmó las afirmaciones de Palantir y los funcionarios de la ciudad de que las intervenciones basadas en datos estaban detrás del descenso temporal de los delitos violentos.
"Es exactamente esta faceta del modelo de la lista 'heat' de Chicago lo que ha expuesto al CPD a una gran cantidad de escrutinio público".
A pesar de que los llamados entraron, los correos electrónicos obtenidos por The Verge indican que el NOPD siguió usando Palantir para hacer cumplir la ley. Palantir rechazó solicitudes reiteradas de comentarios, pero los correos electrónicos también muestran que la empresa estaba al tanto de los riesgos potenciales planteados por los algoritmos de vigilancia predictiva y la publicidad negativa que viene con ellos. El 23 de mayo de 2016, la ingeniero de libertades civiles de Palantir, Courtney Bowman, respondió a una solicitud del analista de delitos NOPD Zach Donnini sobre si Palantir podría ayudar a generar clasificaciones numéricas para el riesgo individual de cometer o ser víctima de un tiroteo.
"Tengo algunas preocupaciones serias sobre la institución de un enfoque de clasificación o puntuación numérica", escribió Bowman. "Es exactamente esta faceta del modelo de lista de" calor "de Chicago la que ha expuesto al CPD a una gran cantidad de escrutinio público", dice el correo electrónico, que enlaza dos artículos que critican el enfoque policial predictivo de Chicago.
"La inminente preocupación es que un algoritmo de puntuación opaco sustituye el barniz de certeza cuantitativa por un juicio más holístico y cualitativo y por culpabilidad humana", escribió Bowman. "Una de las virtudes perdurables del trabajo de SNA que hemos realizado hasta la fecha es que hemos mantenido a los analistas humanos al tanto para garantizar que las redes se exploren y se analicen de una manera que supere la prueba directa".
Independientemente de la sostenibilidad de la reducción de homicidios de Nueva Orleans, Palantir usó su trabajo con el NOPD para solicitar grandes contratos con otras ciudades estadounidenses. Más tarde, la compañía ganó contratos lucrativos para programas predictivos con gobiernos extranjeros.
Según los correos electrónicos obtenidos por The Verge, el personal de marketing de Palantir contactó primero con el Departamento de Policía de Chicago a fines de 2013 sobre la posibilidad de vender un paquete de vigilancia predictiva basado en el trabajo de la empresa en Nueva Orleans, y finalmente pactar un precio de $ 3 millones. A través de una serie de subvenciones federales otorgadas a CPD a partir de 2009, la policía de Chicago y académicos del Instituto de Tecnología de Illinois ya habían creado su propio programa de previsión del delito que asignaba un puntaje de riesgo a las personas en base a datos criminales e historias de redes sociales.
El 19 de agosto de 2014, Katie Laidlaw, ejecutiva de marketing de Palantir, envió un correo electrónico al comandante de la policía de Chicago, Jonathan Lewin. "Me gustaría continuar con la conexión con el Superintendente McCarthy, específicamente para enmarcar el posible compromiso de Palantir con nuestros resultados probados en el apoyo a la reducción de homicidios en Nueva Orleans", escribió Laidlaw.
Palantir usó su trabajo con NOPD para solicitar grandes contratos
Los correos electrónicos también muestran que la Policía de Chicago esperaba recibir dinero de la subvención del Departamento de Seguridad Nacional para financiar la adquisición del software Palantir. However, the Chicago Police Department never piloted or purchased Palantir’s software.
Commander Lewin, who is in charge of Chicago’s “heat list” model of predictive policing and who was on the receiving end of Katie Laidlaw’s sales pitch for Palantir, said in an interview that he was aware of Palantir’s work with other law enforcement agencies but never approved either a test run or purchase of Palantir software.
Though Palantir did not succeed in selling its New Orleans-tested tools to Chicago Police, the data-mining company has successfully sold forecasting products to foreign security services.
In 2016, the Danish national police and intelligence services signed an 84-month contract with Palantir — reported in the Danish press to have been worth between $14.8 and $41.4 million — for a predictive technology package meant to identify potential terrorists. According to procurement documents, the program uses law enforcement data like license plate reader records, CCTV video, and police reports to make predictions about individuals’ likelihood to commit terrorism. Denmark’s national legislature had to pass an exemption to the European Union’s data protection regulations in order to purchase Palantir’s software.
Denmark had to pass an exemption to the European Union’s data protection regulations in order to purchase Palantir’s software
Prior to the 2016 contract with Denmark, Palantir Technologies’ reported work with law enforcement never mentioned forecasting or prediction capabilities.
Last year, the liberal Israeli newspaper Haaretz reported that Israel’s security services used analytics systems that scraped social media and other data to predict potential “lone-wolf” attackers from Palestinian communities in the West Bank, and that Palantir was one of only two technology companies to provide predictive intelligence systems to Israeli security organizations. The New Orleans project is the first reported instance of Palantir using social media data as a part of the company’s social network analysis.
“I’m not surprised to find out that people are being detained abroad using that information,” said New Orleans council president Jason Williams, pointing out the differences between the legal systems of Israel and the United States. “My concern is, the use of technology to get around the Constitution — that is not something that I would want to see in the United States.”
Around the country, cities like New York are weighing legislation about how to oversee the algorithms government agencies use to make decisions. These debates have yet to begin in New Orleans, where the city’s intractable crime rate takes up much of the oxygen in public discourse. However, the secrecy of Palantir’s relationship with NOPD raises red flags to outside observers and prompts questions about how the company’s algorithms are being used.
William Isaac, the Michigan State researcher who has analyzed predictive policing systems for bias, said he has long had suspicions that Palantir engaged in some sort of individual forecasting program. “They had only publicly acknowledged the extent to which their technology is data deconfliction and visualization,” Isaac said.
After being walked through the documentation of Palantir’s New Orleans project, Isaac said the program was remarkably similar to Chicago’s individual “heat list” model, which a RAND Corporation study found had no impact on violent crime and was overwhelmingly composed of young African-American and Latino men with extensive law enforcement contact.
“The same flaws that were in the Chicago predictive program are going to be amplified in New Orleans’ data set.”
“If you’re trying to predict anything, you need to have some representation across the universe that you’re trying to predict. If you’re trying to predict crime, you need to have positive and negative examples for every possible offense,” Isaac said. Police departments tend to have good data about communities where they are present but little data about communities where they do not patrol as vigorously — which tend to be affluent and white.
“The same flaws that were in the Chicago predictive program are going to be amplified in New Orleans’ data set,” Isaac said.
The secrecy surrounding the NOPD program also raises questions about whether defendants have been given evidence they have a right to view. Sarah St. Vincent, a researcher at Human Rights Watch, recently published an 18-month investigation into parallel construction, or the practice of law enforcement concealing evidence gathered from surveillance activity. In an interview, St. Vincent said that law enforcement withholding intelligence gathering or analysis like New Orleans’ predictive policing work effectively kneecaps the checks and balances of the criminal justice system. At the Cato Institute’s 2017 Surveillance Conference in December, St. Vincent raised concerns about why information garnered from predictive policing systems was not appearing in criminal indictments or complaints.
“It’s the role of the judge to evaluate whether what the government did in this case was legal,” St. Vincent said of the New Orleans program. “I do think defense attorneys would be right to be concerned about the use of programs that might be inaccurate, discriminatory, or drawing from unconstitutional data.”
If Palantir’s partnership with New Orleans had been public, the issues of legality, transparency, and propriety could have been hashed out in a public forum during an informed discussion with legislators, law enforcement, the company, and the public. For six years, that never happened.