Kraig Lewis decided at a young age that he valued education: “I grew up in a rough neighborhood, but education was always important to me and my mom…I went to private school…I just wanted to be able to work like my mom. My mom was a registered nurse, and she was able to take care of me.” Kraig was friends with kids at his private school and kids in his largely low-income neighborhood in the North Bronx, even though those two groups rarely overlapped. Like any kid, Kraig didn’t really give much thought to the fact that he had two different sets of friends—they were just…his friends. 

Kraig hoped to become a lawyer someday, and his first step towards that goal was attending college in Connecticut. Kraig started building the life he wanted for his family while he was in college—when he was 18, his son was born, and he lived close to his college with his girlfriend and son.

After obtaining his college degree, Kraig interned at a law firm, also working at Panera Bread to support himself and his family. Still focused on his goal of becoming a lawyer, Kraig visited a law school to learn more about the process. While there, someone told Kraig he looked like a “thug.” Understandably, the insulting and ignorant comment made Kraig feel unwelcome; it was also discouraging. Despite Kraig’s hard work and law school dreams, he decided that he needed to retool: “So, I tell myself I’m going to get my Master’s [degree] and come back.” With that, Kraig returned to Connecticut for graduate school, where he spent two years working towards his Master’s in Business Administration (MBA).  

In April of 2016, Kraig was wrapping up his final semester of graduate school. He was thriving in his school community, and his dreams of studying law finally seemed within reach. Except, unbeknownst to Kraig, federal law enforcement had a different plan. Based entirely on incorrect information, federal law enforcement officers were going to rip Kraig’s dreams to shreds.

Based entirely on incorrect information, federal law enforcement officers were going to rip Kraig’s dreams to shreds.  

In the early hours of April 27th, 2016, while Kraig’s then-six-year-old son was still asleep, federal law enforcement officers, in full SWAT gear, arrested Kraig in his Connecticut home. This is how he describes his experience: 

When I was hit by the feds, I didn’t really know what was going on…I had to learn federal law on the fly…. It was an out-of-body experience for me. I didn’t know how long I was going to [be away]; I just knew I wasn’t going to cooperate…. [All I kept thinking was] I’ll pray and use this injustice when I come home to help people who…didn’t get to go to school…because if I could be here [after private school, college, and graduate school] and end up sitting where El Chapo was sitting, imagine what could happen to some kids in the Bronx.

Kraig was right—the same thing was happening in the Bronx that very same morning, to his childhood friends. Kraig and his childhood friends were arrested in the largest gang raid New York City’s history, now known as “The Bronx 120 Raid.” The raid was the result of sweeping federal RICO charges enabled by flawed law enforcement data and surveillance tools.

Amazon, Microsoft, and other tech companies are the enablers behind the technology that police officers use to label (often inaccurately) young Black and Brown men, especially those living (currently or even formerly, as in Kraig’s case) in low-income communities, as “gang members.” Thanks to these tools, police can surveil and target these young people and their acquaintances, yet the companies that develop them are not held accountable. Through the use of nondisclosure agreements (NDAs) and other tactics, these companies have even managed to keep their identities secret: their names rarely appear on vendor contracts with law enforcement, rendering them virtually anonymous and impervious, without extensive investigative work, to the harmful consequences of their products.

Congress enacted the federal criminal Racketeer Influenced and Corrupt Organizations Act (also known as 18 U.S.C. §§ 1961–1968 or “RICO”) in 1970 in response to constituents’ growing concerns about the threat of La Cosa Nostra (also known as the “Mafia”), a complex enterprise involving many people who were engaged in organized criminal activities throughout the United States, and which infiltrated and threatened legitimate businesses and labor unions. Then, in the 1980s, the “War on Drugs” incentivized elected officials to shift their focus from white-collar and Mafia forms of organized crime to alleged drug trafficking and violence in low-income urban neighborhoods. As the “War on Drugs” garnered attention, the Supreme Court revisited and expanded its definition of a RICO enterprise in the 1981 case of United States v. Turkette. The result advantaged prosecutors in targeting predominantly low-income Black and Brown people perceived to be threatening at the height of the crack boom.

Since Turkette, prosecutors have further expanded their use of the RICO statute to encompass even more loosely defined, still predominantly Black and Brown, “street gangs,” such as the Bloods, Crips, MS-13, and even small neighborhood “crews.” Indeed, in recent years, few have felt the impact of RICO more than young Black and Brown men in low-income communities—many of whom were not “gang members” at all. 

Empowered by the expansive authority conferred by access to the RICO Act, law enforcement officers rely on gang databases to target possible suspects for RICO investigations, arrests, and prosecutions. These databases are notoriously inaccurate. Gang databases demonstrably maintain an overrepresentation of young people of color—with entries based on factors as innocuous as wearing clothing of a particular color or displaying certain sports team logos (including, for example, the Brooklyn Nets), interacting with previously-entered individuals, or simply standing in locations designated “gang territory” by law enforcement officers.

The overrepresentation of young people of color in gang databases inevitably drives the overrepresentation of Black and Brown men in gang raids and the ensuing RICO prosecutions, plea bargains, and lengthy incarceration sentences.

The overrepresentation of young people of color in gang databases inevitably drives the overrepresentation of Black and Brown men in gang raids and the ensuing RICO prosecutions, plea bargains, and lengthy incarceration sentences. The result is that the dragnets created by RICO street gang prosecutions entangle innocent young people who neither contribute to nor play a role in complex criminal conspiracies. And yet, despite that reality, federal prosecutors obtain RICO convictions, pleas, and sentences with relative ease.

One of the arrested alleged gang members outside of the 43rd Precinct located at 900 Fteley Avenue in the Bronx on Wednesday, April 27, 2016. Over a hundred suspected gang members were arrested after a gang bust in the Bronx. (Anthony DelMundo/New York Daily News)
Permission for use granted by New York Daily News. Image was originally used in the following slideshow: 120 Bronx Gang Members Indicted in Largest Takedown in NYC History: Look Back at 87 Arrests in One Day, N.Y. Daily News (Feb. 25. 2019).

On the same April morning when federal law enforcement officers ripped Kraig from his Connecticut home and his sleeping six-year-old son, helicopters whirred and sirens blared in the North Bronx. Over 700 officers, with no-knock warrants and guns drawn, broke into apartments and woke the sleeping families inside. Young children awoke, paralyzed with fear, to the sight of shouting strangers, dressed in riot gear and carrying weapons, stomping through their homes; teenaged girls were forced out of their beds, some in just their underwear, by screaming male officers who would not allow them to retrieve clothing from their bedrooms; mothers and grandmothers tried to comfort their families and calmly answer questions from officers demanding to know if they were hiding anyone else, while their husbands, sons, and grandsons lay face down, in their boxer shorts, with handcuffs around their wrists. 21-year-old Geovanni Martin tried to escape through a window and fell to his death (in a 2020 documentary entitled “Raided, Part 2” his best friend explains that she believes Geovanni feared for his life because, in 2012, police shot their unarmed classmate, Ramarley Graham, in his home). In short, the raid was a living nightmare.  

In short, the raid was a living nightmare. 

The U.S. Attorney for the Southern District of New York (SDNY), the New York Police Department (NYPD), U.S. Immigration and Customs Enforcement (ICE), Homeland Security Investigations (HSI), U.S. Drug Enforcement Administration (DEA), and the U.S. Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) SWAT teams rounded up and arrested 88 alleged members of two so-called “rival street gangs” in what later became known as the largest “gang raid” in New York City history. In total, 120 people were indicted; those who were not among the 88 arrested during the raid were already serving time for other convictions.

Perp-walk images of young Black and Brown men dominated the news—pictures of young men, in pajamas and handcuffs, trying to cover their faces as police loaded them into vans. The majority of the arrestees were in their late teens and early twenties, with their entire lives ahead of them. Even though these young men were still legally innocent, the U.S. Attorney’s Office published a long list of their names in its press release about the raid.

The SDNY indictments, formally entitled United States v. Burrell and United States v. Parrish, alleged racketeering conspiracy, narcotics conspiracy, narcotics distribution, and firearms charges. It was later reported by two professors, Babe Howell (of CUNY Law) and Priscilla Bustamante (of Baruch), that dozens of the young people arrested weren’t actually in gangs; rather, many of the arrests were based on whom the individuals knew, rather than what they actually did, i.e., guilt by association:

One of the most startling revelations of the review of the Bronx 120 prosecutions is that half of those swept up in the largest gang raid in the history of New York were not affirmatively alleged to be members of either of the two rival gangs allegedly targeted by the mass indictments. The prosecutor’s sentencing submissions…affirmatively state that 34 of those subjected to the raid and arrested as part of the RICO case were not gang members. An additional 17 individuals are characterized as “associates of” or “associated with” the two rival gangs…. Thus, 51 of the defendants swept up in “the largest gang takedown in New York City history” were affirmatively not alleged to be gang members. For another 13 there is no clear allegation relating to gang membership. Their dispositions suggest they were not gang members. 

The Bronx 120 Raid destroyed lives, families, and communities: 80 of the young men and teens arrested were not charged with violent offenses. Over 100 took plea deals, regardless of their actual guilt. Two of the five murders cited in the indictment had already been solved before the raid even took place. Over 90% of the people arrested were U.S. citizens, but ICE nonetheless played a significant role in the raid, due to its “transnational gang program” (a program that can be triggered by just one non-citizen in a group of acquaintances allegedly participating in some form of illegal activity). The New York City Housing Authority (NYCHA) responded to the raid by investigating families in apartments where someone had been arrested; at least one mother whose child was named among the Bronx 120 defendants was evicted and became homeless as a result of the NYCHA investigations.  

None of this mattered—not at the time of the raid and not thereafter. Law enforcement insisted that these young men were dangerous—guilty by association—and urged that the public safety benefit of the militarized raid sufficiently outweighed the costs of the raid for those affected.  

Alice Speri, a reporter for The Intercept, spoke with families affected by the raid. They told a very different story: “[I]n community meetings and private conversations, local residents have expressed rage over the raids and said that when they asked for safer neighborhoods, they didn’t mean officers in riot gear breaking into their apartments in the middle of the night and taking away their children.”  

Similarly, Professor Babe Howell questioned, in an article she wrote after the Bronx 120 Raid, why law enforcement decided against simply working directly with community members instead of relying upon secret surveillance tactics and information gathering technologies:

First, one may question the wisdom of watching, listening, spying, waiting and then using conspiracy charges to link dozens of young people to offenses committed by others instead of intervening to defuse the rivalry. Second, one may wonder how a military-style raid to accomplish regular law enforcement goals affects police-community relations. Having obtained the indictment and surveilled the individuals for years, why enter their homes wearing bulletproof vests, with firearms drawn, pointing weapons at family members, [with] helicopters overhead?

How did a massive “precision policing” so-called gang raid, carried out by hundreds of officers from half a dozen federal law enforcement agencies, arrest and publicly brand as “dangerous criminals” so many non-gang members? How did a highly coordinated action, with massive resources at its disposal, place innocent young people on a path to that basketball court behind razor wire, or worse—adult prison? And why were so many of these young people sent to prison under the extreme threat of the RICO Act? In order to answer those questions, we’ll have to examine the flawed bedrock of so-called precision policing: gang databases and police surveillance technology.

Caption Required by NY Daily News – Nearly 100 Bronx gang members were arrested early Wednesday, April 27, 2016 as police and federal authorities conducted a massive gang sweep at the Eastchester Gardens houses and other locations in the Bronx. (Anthony Delmundo/New York Daily News)
Permission for use granted by New York Daily News. Image was originally used in: 120 Bronx Gang Members Indicted in Largest Takedown in NYC History: Look Back at 87 Arrests in One Day, N.Y. Daily News (Feb. 25. 2019).

Once upon a time, it would have been reasonable to assume that law enforcement officers understood the investigative information-tracking tools upon which they relied. However, that is no longer the reality. In fact, the tech companies that create and sell sophisticated investigative and informational technologies to law enforcement can more reasonably assume that their customers do not fully understand their products and are, therefore, unlikely or unable to disclose information about them to (1) elected officials seeking to regulate them, (2) judges approving warrants based on them and considering the risk tech companies pose to constitutional rights, and (3) competitors within the tech industry.

Federal, state, and local law enforcement entities have become major consumers of surveillance and informational technologies that private sector corporations control; thus, these corporations have come to have an outsized influence on law enforcement. Private sector companies insulate themselves from public and judicial scrutiny through things such as NDAs and “trade secrets” claims, so their influence on law enforcement through the technologies they sell often goes unnoticed.  

In order to continually increase market share and shareholder profits, private sector tech companies must ensure that law enforcement becomes and remains reliant on their products.  Therefore, private tech companies have perverse incentives—in some sense, they benefit from the continuance of crime and violence, not from solving or preventing it. The surveillance and informational tools that these companies sell appear to make the jobs of law enforcement easier and more efficient, yet the results are more complicated than simply greater efficiency.

As Professor Elizabeth E. Joh (of the U.C. Davis School of Law) has cautioned, “[t]he companies which provide these technologies act out of private self-interest, but their decisions have considerable public impact. The harms of this private influence include the distortion of Fourth Amendment law, the undermining of accountability by design, and the erosion of transparency norms.” To truly understand the havoc law enforcement can wreak on the most vulnerable communities, we must understand the role of private sector companies, their products, and the ways in which they have leveraged legal and corporate tools to insulate themselves from scrutiny and accountability for the ways those products are being used.

Race and racism are endemic to risk assessment and related algorithms used by law enforcement, the courts, and correctional entities. Indeed, the notion of risk in this context is essentially a proxy for race.

Race and racism are endemic to risk assessment and related algorithms used by law enforcement, the courts, and correctional entities. Indeed, the notion of risk in this context is essentially a proxy for race. Historically, these algorithms explicitly included race and ethnicity as factors, and being anything but white counted against anyone who was evaluated under the algorithms. Once race became a protected class for civil rights purposes, the algorithms reoriented their focus towards prior criminal history instead; however, given the history of systemic racism in the U.S. and its common manifestation in the hyper-criminalization of Black and Brown people, prior criminal history essentially became a proxy for race.  

These invasive technologies threaten rights and liberties, broadly but that effect is exacerbated in low-income communities of color because they tend to be the most heavily policed, regardless of the actual distribution of crime throughout the United States. 

People do not know when or for what reason their names have been added to a gang database. Even if they learn of their inclusion, they lack recourse.

Antonio Reynoso (New York City Council Member representing parts of Brooklyn), and Taylonn Murphy Sr. (a Harlem father-turned-activist whose daughter, Tayshana Murphy, was killed in 2011, and whose son, Taylonn Murphy Jr., was one of the 103 people arrested by the NYPD and FBI during a 2014 “gang raid”) have explained the impact of gang databases:

Inclusion in the database can mean intensive surveillance, police harassment, overcharging, increased bail, risk of deportation and prejudicial treatment in court: a separate track of justice based on an allegation that doesn’t even have to be proven. Despite claims that the database isn’t being shared, the NYPD has coordinated “gang takedowns” with federal agencies.

This approach cannot be reformed. The NYPD has a long history of ignoring and subverting reforms. And there is no “better” database. The database—like stop-and-frisk before it—doesn’t make us safer; it has been fully operational amid rises in shootings and violence.

Gang databases in cities such as Boston, Chicago, the DC-Maryland-Virginia region, Los Angeles, and New York have been criticized for serious inaccuracies. Josmar Trujillo (a journalist and community organizer) and Professor Alex S. Vitale (of Brooklyn College) illustrate in their 2019 report the weak foundations upon which law enforcement officers select people for inclusion in gang databases:

The NYPD admits to categorizing local ‘crews’—smaller, more local and less formal groupings—alongside gangs within its database. Like gangs, ‘crews’ have no consensus definition. Therefore, the NYPD gang database [is] a database of people that police believe to be grouped together. There is no requirement of a criminal conviction, much less a violent conviction, to [be] added to the database.

Disconcerting revelations about gang database inaccuracies abound. For example, a 2016 state audit revealed that California’s state gang database (CalGang) contained “scant documentation for nearly one in five of the 100,000 database entries. Many people remained in the system long after they should have been purged . . . with no documented gang activity. [Also,] CalGang inexplicably contained reports on several dozen purported gang members [who were] less than a year old.” Journalist Mike Dumke has reported in ProPublica that as recently as 2017, the Chicago Police Department gang database included over 100 people who exceeded 75 years of age; information about the same database in 2018 revealed that 13 “gang members” were approximately 118 years old and two were 132 years old.

Suspected “gang members” aren’t the only people populating the impenetrable gang databases; law enforcement also tracks suspected “gang associates,” e.g., people who have dated suspected “gang members,” or even one remove—people who have close relationships with other people who formerly dated “gang members.” Clothing can matter, too. Journalist John Blitzer wrote an article in The New Yorker describing the experience of a young man named Jorge who lived on Long Island. Police arrested Jorge for three reasons: (1) a police officer at his school saw him wearing a Brooklyn Nets hat—a local sports team logo that law enforcement officers associate with MS-13 membership; (2) police saw him perform in a special handshake with a friend—although police did not provide any detail about this in their reports; and (3) “he was observed in the company of two people on ICE’s radar…” including a gang database designee and Jorge’s own girlfriend, who previously had dated, but was later kidnapped and assaulted by, someone who notoriously claimed MS-13 membership.  

ICE requires satisfaction of at least two criteria, from a lengthy list, in order to label someone a “gang member.” Examples of ICE’s criteria include: “frequenting an area notorious for gangs” and “wearing gang apparel” (like Jorge). But, as Jorge’s lawyer, Alexandra Lampert, pointed out, those criteria mean that “[a]ny student at [Jorge’s] High School would be at risk of arrest just for interacting with other students and wearing clothing representing local sports teams.” Immigration experts Laila L. Hlass and Rachel Prandini have reported that ICE has even added a child to its gang database for wearing a blue shirt that was actually just part of the child’s mandatory school uniform. In New York City, acts as innocuous as lending a cell phone, wearing a shirt of a particular color while standing on a particular street corner, or posting an off-hand reference to an alleged gang on social media can land someone in a gang database and forever after implicate that person for “furtherance of a conspiracy.”  

ICE has even added a child to its gang database for wearing a blue shirt that was actually just part of the child’s mandatory school uniform.

Inclusion in a gang database tends to serve as a precursor to highly complex prosecution under some of the broadest criminal statutes in the United States criminal legal system that impose some of the harshest prison sentences on those convicted. According to activist Josmar Trujillo, New York City officially considers most low-income housing to be “gang territory,” which means that the housing complexes in which people live can serve as criteria for gang database inclusion. Associating with “known gang members” is another criterion law enforcement uses to label people as “gang members,” but if everyone who lives in a particular housing complex hangs out with their friends in that complex (who are also likely their classmates in the local school), then everyone can easily end up in a gang database if just one person is already designated by law enforcement.

Fundamentally, gang databases do not make people safer. They enable police to over-criminalize and harass even more people than they otherwise might in communities that often have long histories of tensions with law enforcement.

Fundamentally, gang databases do not make people safer. They enable police to over-criminalize and harass even more people than they otherwise might in communities that often have long histories of tensions with law enforcement. Furthermore, since these gang databases serve as a precursor to sweeping RICO indictments, they increase the likelihood that the people included in gang databases will end up incarcerated. The traumatic, dangerous, and inhumane conditions endured by incarcerated people can heighten their likelihood of violence and gang membership—if only due to a need for protection while in custody. Moreover, when people are released from incarceration, they have felony records that make it harder for them to find housing and employment, so the entire practice sets them up to fail. None of this improves public safety.

Trujillo and Vitale have highlighted the importance of social media in arbitrary gang database determinations by law enforcement:

Police interpretation, or perhaps willful misinterpretation, of gang admission on social media can include emojis, hashtags, or other forms of communication. There is also the question of how police can authenticate who is posting or operating a social media account. Making matters worse, the use of social media posts as a way to authenticate gang membership significantly expands an already questionable process by turning the internet into a virtual police precinct. 

Gang database inclusion serves as a predicate for surveillance. Journalist and attorney Keegan Stephan has asserted that law enforcement has the power to subject people who are manually added to gang databases to ongoing social media monitoring and other forms of surveillance and can then add the social media contacts of those people (e.g., “friends” or “followers”) to the same database. Law enforcement can subsequently use information obtained through its exceedingly broad social media surveillance to construct criminal cases. Notably, even though people are added to gang databases manually, the surveillance technology law enforcement officers use can assist those monitoring social media with quickly compiling substantial biographical and personal information when a person is selected for further investigation.

Gang database inclusion serves as a predicate for surveillance.

Attorney Naz Ahmed, who works with the CUNY Creating Law Enforcement Accountability & Responsibility Project (CLEAR), which serves individuals who are either being surveilled by the FBI or approached to help the FBI surveil Muslim communities, has emphasized the overlap of surveillance and racial profiling practices identified by CLEAR and the practices that precipitated the Bronx 120 raid. In a discussion with journalist Max Rivlin-Nadler, Mr. Ahmed cautioned that “[g]ang raids and Muslim surveillance are no different. The NYPD was surveilling these kids when they were twelve. The FBI does the same thing to Muslim communities using the same or similar electronic surveillance programs that look at the online activities of kids and say, ‘Oh, you’re going to be a terrorist.’”

Palantir Technologies (Military-Grade Surveillance)

In the case of the Bronx 120 Raid, a data analytics corporation called Palantir Technologies, which was co-founded by Trump transition team advisor, billionaire, and conservative campaigns funder Peter Thiel, often works with federal law enforcement agencies. For example, ICE contracted with Palantir for the use of its FALCON software program, which compiles and manages data from government agencies, such as the Federal Bureau of Investigation (FBI) and branches of the military.

General Dynamics Information Technology and Qbase-McNeil (Military-Grade Surveillance)

General Dynamics Information Technology (which works with ICE and the Department of Defense) owns what used to be known as Qbase-McNeil Integrated Solutions. In an article in The Appeal, journalist Simon Davis-Cohen described Qbase-McNeil as “a company that supports ICE ‘with mission-critical IT service support and program management both domestically and internationally,’ to coordinate laptops and a virtual private network for government agents. Qbase-McNeil also assists the Defense Information Systems Agency…” and other government agencies.

Dunami – by Q-Tel and Oracle

According to Chicago’s Lucy Parsons Labs (LPL), the Chicago Police Department (CPD) uses social media monitoring software called Dunami as part of its anti-gang efforts. From 2014-2018, CPD partnered with over two dozen schools in the South and West Sides of Chicago to conduct social media monitoring focused on identifying potentially gang-involved students. (Note: CPD previously used a gang database called the Chicago Law Enforcement Analysis and Reporting System, or “CLEAR,” which it developed in partnership with computer technology giant Oracle.) 

Dunami (made by PathAR LLC) is an In-Q-Tel-funded software that is used for social media intelligence and analytics. Notably, In-Q-Tel is apparently “the venture capital arm of the CIA.” LPL highlights that “Dunami has been used by FBI and CBP ‘to determine networks of association, centers of influence and potential signs of radicalization.’ Dunami has been used by CPD and Chicago Public Schools to monitor for ‘signs of violence and gang membership’ among students.” Through Freedom of Information Act (FOIA) requests, LPL gained access to financial records showing that “CPD spent more than $1.25 million on Dunami…between 2014 and 2018,” in addition to hundreds of thousands of dollars on similar products.

Law enforcement often finds social media material threatening when it is actually only portraying toughness in order to deter violence (or, in the case of some drill music artists, to make money in a legal manner).

Social media surveillance of this kind has received extensive criticism. The monitoring algorithms often misunderstand or misinterpret complex human relationships, jokes, slang, etc., and they tend to produce frequent false positive alerts as a result. Professor Forrest Stuart has discovered through his sociological research that law enforcement often finds social media material threatening when it is actually only portraying toughness in order to deter violence (or, in the case of some drill music artists, to make money in a legal manner). Indeed, law enforcement misinterpretations of social media posts can cause individuals posting and anyone included in the posts to be added indiscriminately to gang databases. Professor Stuart’s 2020 book, Ballad of the Bullet explores these tensions in depth.

Microsoft, Azure, and the Domain Awareness System

Microsoft has been especially evasive about its ties with law enforcement agencies; however, it has been a major player in this space for many years. Microsoft is the creator of the Domain Awareness System, a mass surveillance tool built for the NYPD (Atlanta later adopted it as well).  Azure, a division of Microsoft, provides a “Government Cloud” platform through which multiple police surveillance vendors can operate their products. Microsoft, largely through Azure, has encouraged law enforcement entities to synchronize their field work with products such as drones and robots. One option Microsoft has developed to this effect is known as the Microsoft Advanced Patrol Platform (MAPP). Several smaller corporations provide components of MAPP. This tool essentially converts police vehicles into surveillance patrol machines.  

Microsoft has been exceptionally successful in evading criticism of its law enforcement ties. When Microsoft employees advocated for the company to cancel its contracts with law enforcement, Microsoft sidestepped those demands by “banning” sales of its facial recognition software to law enforcement entities while continuing to develop and sell numerous other law enforcement products. Microsoft and its smaller partners evade public scrutiny in another way, too: the services that host third parties on the Azure cloud are not considered public information, nor is the information that resides there.

The secrecy tactics surrounding the Stingray technology serve as a powerful example of the lack of transparency this largely unregulated industry enjoys. Stingrays are international mobile subscriber identity (IMSI) catchers—they are also known as cell-site simulators. Every user of a mobile telecommunication network has a unique identification number; basically, the devices impersonate cell towers so that phones connect to them instead, allowing the devices to track the connected phones. Author and journalist Jon Fasman describes the secrecy surrounding the use of Stingrays in his 2021 book, We See It All:

Few pieces of police technology are as hidden from public scrutiny as cell-cite simulators. Agencies that use them are often required to sign nondisclosure agreements (NDAs). Wired reported that federal marshals raided the Sarasota Police Department to prevent it from turning over documents related to cell-cite simulator use to the ACLU of Florida.  Prosecutors in some jurisdictions have reached settlements to avoid having to disclose details of cell-cite simulator use in open court…. 

The ACLU of Delaware obtained an extraordinary NDA and cover letter from the Harris Corporation, which makes Stingrays, to a detective with the Delaware State Police, who signed it in 2010. The letter warns that ‘only officers with arrest authority are permitted to use them or have knowledge of how they work…. [They] should not be discussed…in any public setting….’ The NDA forbids police from disseminating information relating to the Stingray and requires them to inform Harris of any ‘court order or judicially mandated disclosure.’ It also purports to give Harris veto power over any disclosure to ‘governmental agencies’—meaning that a state police chief, say, could only discuss his force’s use of Stingrays with his state’s legislature or governor if Harris approved.

Cell-cite simulators do not discriminate—they pull data from all phones in their orbit, including phones in private places such as homes and doctors’ offices.

Because so much of what happens with gang databases occurs in secret, and in an unregulated environment, gang databases generally avoid both public and judicial scrutiny. Consequently, groups seeking information about them have resorted to transparency litigation, such as proceedings to compel records regarding gang database demographic data and inclusion criteria. However, this is a partial solution at best. Professor Hannah Block-Wehba (of Texas A&M University School of Law) summarizes this issue in her article, entitled Visible Policing: Technology, Transparency, and Democratic Control:

The informational dynamics of modern policing technologies grow even more complex when private sector vendors are involved…. [L]aw enforcement agencies often begin new surveillance programs and acquire surveillance tools without disclosing those activities to city council or other oversight agencies, let alone the public. As a result, surveillance and other new policing programs are often initiated without any kind of public input. A web of nondisclosure agreements, some of which purport to override state open records obligations, often surrounds policing technology supplied by private sector vendors.

Professor Block-Wehba proceeds to discuss how prosecutorial practices and local court rules further obscure important information about “digital search warrants and surveillance orders,” keeping that information from the public under formal and informal “presumptions of secrecy.”

Private tech companies evade scrutiny, and sharing details about their products, by asserting “trade secret” claims—even when the party seeking information is a criminal defendant submitting the request within the confines of a protective order. Ultimately, as experts in the law and technology field have cautioned, the intellectual property interests of tech companies inevitably come into tension with the interests of defendants hoping to obtain evidence. When that tension arises, corporate interests almost always prevail. Importantly, that prevalence refers to situations in which the defendant has some knowledge about the technologies that affected a charging decision by law enforcement. In the vast majority, technologies are so concealed by NDAs and other measures that criminal defendants, counsel, and judges never even have an opportunity to consider the merits of access to the information they contain.  

In the case of surveillance technologies, which might arguably constitute searches under the Fourth Amendment, these technologies often exist outside the limits of warrant requirements and Fourth Amendment jurisprudence; if we add NDAs governing what law enforcement can even disclose about surveillance on top of that, judges can never even reach the Fourth Amendment question—perpetuating the total lack of restrictions on surveillance by law enforcement that corporations seek to preserve. These tech companies are competing with each other. The NDAs are safeguards the companies use to silence their customers and avoid criticism, but the companies have no desire to disclose information about their methodologies anyway. To do so would benefit their competitors. In the case of surveillance companies, in particular, they are inherently competing to collect the most information about the greatest number of people—that is not going to incentivize transparency, even in the face of criticism.

These technologies often exist outside the limits of warrant requirements and Fourth Amendment jurisprudence; if we add NDAs governing what law enforcement can even disclose about surveillance on top of that, judges can never even reach the Fourth Amendment question—perpetuating the total lack of restrictions on surveillance by law enforcement that corporations seek to preserve.

As the National Association of Criminal Defense Lawyers Task Force on Predictive Policing has noted, “[t]he lack of transparency surrounding data-driven policing and the industry’s fundamental business model are interwoven. Private companies compete…. [S]ome companies even provide technology to police departments initially for free, with the goal of selling departments on the need to continue using the technology…. [M]any departments themselves often possess little to no insight into the inner workings of the systems they employ and lack incentive to do so without explicit transparency measures in place….” Professor Rashida Richardson (of Rutgers Law School and the Rutgers Institute for Information Policy and the Law) highlighted the following for that same task force:

[A] more public and transparent and accountable process would also just look bad for [police departments] everywhere, when they’re kind of like, ‘I don’t actually know how this system works. We are just using it because we got it for free, and we thought it would help.’ When you don’t have data to prove its efficacy as backup, which they often don’t, then it raises a lot of good governance questions: like why are you using these systems, why are public tax dollars being used for this?

Ultimately, these private tech companies work to build police dependence on their products and avoid transparency and accountability as they do so.

Once the technologies described above have facilitated mass arrests, prosecutors are able to leverage their other advantages over defendants. One such advantage flows directly from the RICO statute. Prosecutors in RICO cases need only meet the low threshold of demonstrating (1) the existence of an “agreement” and (2) an intent, or apparent intent, to commit some sort of crime. Thus, a small group of young acquaintances could posture among themselves about stealing from a local store, even if only half-serious or as a joke, but if one of them, acting on his own, looks up the hours or address of that store online at some point (which courts perceive to be an overt act towards committing the intended offense), a prosecutor could actually prove a conspiracy charge against the entire group of acquaintances.

Prosecutors and courts interpret the RICO Act broadly in street gang cases. RICO convictions or plea bargains inflict severe sentences by design (including often sentence enhancements for the activities of other people whom a defendant barely knows or might not know at all). Consequently, young Black and Brown men, regardless of guilt, often decide to plead guilty in order to avoid harsher RICO sentences, thereby allowing RICO prosecutions to circumvent judicial and public oversight.  

The RICO statute requires the government to establish at least two predicate acts (e.g., drug dealing, bribery, theft, gambling) in order to prove a pattern of racketeering activity, “the last of which [must have] occurred within ten years (excluding any period of imprisonment) after the commission of a prior act of racketeering activity.”

Racketeering activities include state-law felonies and violations of federal law. However, because federal RICO cases are subject to fewer rules than most state cases, prosecutors have a much easier time winning—with far less proof. Federal RICO defendants, identified through flawed databases and surveillance, who were previously acquitted of a state law violation—because prosecutors lacked sufficient proof—could nonetheless later be convicted of a RICO charge on the exact same proof because it would be sufficient to prove that charge to be a predicate act for RICO. Some Bronx 120 defendants experienced this nightmare first-hand: after they had been acquitted of state law charges, federal prosecutors successfully elevated those mere allegations to predicate acts for the purpose of establishing a RICO-qualifying pattern of racketeering activity.

Some Bronx 120 defendants experienced this nightmare first-hand: after they had been acquitted of state law charges, federal prosecutors successfully elevated those mere allegations to predicate acts for the purpose of establishing a RICO-qualifying pattern of racketeering activity.

Most violent crimes are handled in the state system. Consequently, and because convictions for violent crimes carry huge consequences, state laws for such criminal charges afford robust protections. Thus, greater procedural and evidentiary protections exist for defendants in state criminal courts than are available in the federal system, which was developed to handle a broader variety of, largely non-violent, charges. This disparity creates a problem when, in federal RICO cases, prosecutors charge defendants for violent predicate crimes, using the laws of the state in which the defendant allegedly committed the crime, but applying only the reduced procedural and evidentiary protections of federal courts when doing so. 

And then there’s the broad RICO definition of an “enterprise.” Under RICO, enterprises needn’t be particularly organized or collaborative in their decision-making. That expands the applicability of the “enterprise” label beyond its original (Mafia-inspired) conception. While RICO requires an enterprise to affect interstate or foreign commerce, that “effect” can be minimal (in fact, courts have held that the use of telephones, cell phones, the U.S. Postal Service, the internet, and even the distribution of condoms manufactured in another state, can be “instrumentalities of interstate commerce” under RICO).   

The combination of this broad definition of “enterprise” with over-inclusive databases and unfettered surveillance leads to far too many tragedies like the one Kraig Lewis endured.

New Yorkers protesting the NYPD gang database as part of the #erasethedatabase movement.
G.A.N.G.S. Coalition (https://gangscoalition.org/)

Some posit that RICO gang prosecutions, and the gang raids that precede them, allow law enforcement to reduce crime, despite their inaccuracies and however they are achieved—that the end justifies the means. The reasoning goes like this: some young men who do not ultimately receive RICO convictions are nonetheless convicted of lesser offenses and taken off the streets, so we’re still catching dangerous people, even if they aren’t ultimately convicted of RICO charges. Even if true, which I do not concede it is, those incarcerations—made possible by erroneous gang databases, frequently-misinterpreted surveillance, and high rates of largely unsupervised plea bargaining—undermine the legitimacy of law enforcement and the criminal legal system writ large. 

The RICO statute lends itself to being misused as described insofar as it only requires law enforcement to tie one or a few individuals (“targets”) to predicate charges in order for RICO to ensnare many other additional people (“affiliates”) who are in any way associated with those targets; the statute does not even require so-called affiliates to have committed any predicate offenses at all. Thus, law enforcement officers can arrest the acquaintances of RICO targets and then proceed to charge them for non-RICO, non-predicate offenses. The gang databases and electronic surveillance mechanisms enable the identification of many acquaintances. If the acquaintances of RICO targets suffer a presumption of guilt by association and are primarily investigated after arrests that are based on flawed information, all without due process protections, the Constitution is subverted.

Others assert that these ends validate the means—in other words, the practice of charging young Black and Brown men as street gang members under RICO must be justifiable because courts ultimately convict many of the people charged, whether for RICO or other offenses. In other words, the wrongful convictions of wrongfully charged defendants represents judicial validation that (1) alleged street crews are complex criminal conspiracies, and (2) gang databases provide a reliable and lawful basis for prosecuting alleged street crews under RICO. But the premise is false in that courts rarely engage in such validation insofar as convictions do not necessarily reflect guilt, judicial oversight, or jury approval.  

In a cascade of calamitous effects, members of some low-income communities have expressed that people fear reporting crime because of the possible unintended consequence that law enforcement will add them, or someone they know, to a gang database. Concerns about how ICE uses gang databases to spur deportation proceedings, and the erosion of trust in law enforcement due to police brutality, exacerbate this phenomenon. Moreover, raiding an entire neighborhood and removing over 100 young men inevitably harms any community; as Mr. Taylonn Murphy explained in reference to the 2014 raid in his community, “[w]hen you take 103 individuals out of any community . . . you absolutely and positively take away a generation. You’re weakening our community. You’re losing all these resources[.]”

“When you take 103 individuals out of any community . . . you absolutely and positively take away a generation. You’re weakening our community. You’re losing all these resources.”

Thus, although residents of low-income urban neighborhoods who live in close proximity to violence have justifiable safety concerns, including sometimes concerns about gang activity, if they cannot safely call for help, our misguided approach to public safety actually leaves them doubly vulnerable. The combination of flawed gang databases and RICO prosecutions have devastating consequences for the safety of the communities they purport to help.

Surveillance, data, and information technologies used by law enforcement and the criminal legal system are unaccountable and invasive. As law enforcement agencies increasingly contract with private corporations to enhance and expand their use of data and information technologies, the need to fully understand and address the resulting issues grows more critical. Transparency is essential, but that must mean more than just the disclosure of highly technical algorithms out of context. These are complex technologies, and they’re constantly evolving.

As in other spheres of our lives, we cannot safely leave scrutiny of these tools to the handful of experts who have the technical expertise to create them and the corporations that strategically sell them to increase market share and shareholder profits—often to the detriment of anyone who might be adversely affected. As Jay Stanley (of the ACLU) told the National Association of Criminal Defense Lawyers Task Force on Predictive Policing, “[Y]ou need the thousand eyeballs…because if you’re just getting a company that’s offering the product telling you that it’s working, it’s not worth a lot…. We need to make [decisions about these technologies] democratically. And we can’t make those decisions if we don’t know what’s going on.”

We cannot safely leave scrutiny of these tools to the handful of experts who have the technical expertise to create them and the corporations that strategically sell them to increase market share and shareholder profits—often to the detriment of anyone who might be adversely affected.

Defense attorneys, their clients, and the general public lack access to critical information about the surveillance and information technologies used to monitor all of us, and to strip some of us, sometimes wrongfully, of freedom. Nobody, including judges and elected officials, can assess the value, and the risks, of these technologies unless we can comprehend them. When someone is arrested and charged based on, for example, predictive policing algorithms, that person should be entitled to know the factors that contributed to law enforcement’s decision to threaten his liberty.

How can defendants successfully defend against information they do not have? How will anyone trust a system that charges people based on secret information over which practically nobody has any oversight? Adding to the multi-faceted power imbalance that results from the partnership between private tech companies and law enforcement, residents of the communities most affected by tech-heavy policing strategies are further marginalized because they have no influence over their use.  

Policing driven by sophisticated data and surveillance technologies has become increasingly popular among law enforcement entities throughout the United States. Those tools have made policing more invasive and aggressive than ever before, and law enforcement entities cite the costly technologies as justification for increased, rather than decreased, funding. Policing in the United States costs over $100 billion each year, and the growing costs for sophisticated surveillance and information technologies threaten to absorb even more resources that could be better spent on other, proven, community-and-public-health-based public safety strategies. 

When someone is arrested and charged based on, for example, predictive policing algorithms, that person should be entitled to know the factors that contributed to law enforcement’s decision to threaten his liberty.

Moreover, these technology-based risk assessment tools are not even delivering as promised. Rather, as Professor Bernard Harcourt (of Columbia Law School) identified in a conversation with Bianca Tylek (the Founder and Executive Director of Worth Rises) these tools are merely predicting “the likelihood of being policed,” which in turn depends on policing. Thus, we’re looking at the wrong data to predict crime—we’re only looking at things like the likelihood that the police will search, surveil, or detain a particular individual, or broadly defined group of individuals. The result is the systemic loss of liberty and rights at the hands of corporations motivated by profits—these corporations are producing and selling products that are used by law enforcement to hold people’s lives in the balance, based on secret ingredients that remain largely unknown and entirely unchecked. In his recent book, Jon Fasman put it this way: “If our civil liberties imperil the profits of tech entrepreneurs and firms that sell surveillance technology, they are free to choose the profits every time.”

When Kraig Lewis was in the midst of his RICO nightmare, he felt his world crashing down: “When it was happening to me, I never thought anybody cared…. I wanted to die…. I just want you to understand—I had the perfect life…. [Eventually, my judge] basically apologized to me…. They let me out after two years. They had no evidence on me. Somebody probably told them something, and after a while they probably wouldn’t have been able to corroborate it. I was in school, and that’s what I was focused on. I could have been a judge, a lawyer…. They really set me way back.” Kraig’s tragic story is just one of the many.

This article only offers only a sampling of the data collection and surveillance partnerships currently in effect between corporations and law enforcement. Some advocates of aggressive policing assert that any tool that helps police catch and incarcerate more “criminals” is a worthwhile investment. However, as Jon Fasman says in his book, “[t]here are many things—such as indefinite detention without charge and repealing habeas corpus –incompatible with free speech and open society that would make the job of law enforcement easier…. It is our responsibility to speak up for ourselves, our civil liberties, and the sort of world we want to see….”

Kraig Lewis is doing just that. He was recently hired by the Legal Aid Society as a community organizer. Kraig remains astounded by how little his lawyer cared about his situation and his innocence. Still Kraig sees this community organizing position as a step, albeit delayed, towards law school. He wants to help people and protect them from the terrors law enforcement officers can inflict on people like him and communities like his own. He understands, personally and deeply, the role of police surveillance—corporately-enabled and incorrect—in destroying young Black and Brown lives, and he refuses to give up his fight against it.

The law has not kept up with these public-private partnerships and the secrecy surrounding them. Police officers use these technologies without warrants because they do not easily fit within the kinds of searches courts have typically recognized as requiring warrants. It is also difficult to challenge the invasive use of something that we don’t know and can’t see. Protections must be sought on multiple fronts: some combination of legislation, regulation, transparency, grassroots efforts, and even, someday, abolition.

From The [F]law:

From The Systemic Justice Journal:

• Austin Nielsen-Reagan, The Profitability of Inhumanity: How Corporate Power Gives Rise to Forced Labor in Privatized Immigration Detention
• Adira Levine, Can’t Get Back Up: The Growth and Toll of Private Probation