TECHNOLOGY &
CRIMINALIZATION

AI and decision-making algorithms are becoming a common feature of tech-driven policing and detention systems. AI technologies such as facial recognition software, predictive policing, and pre-trial risk assessment algorithms perpetuate criminalization through racial and gender bias and have become what Joy Buolamwini of the Algorithmic Justice League calls the “coded gaze.” These algorithms are biased because these types of automated systems are designed within an already discriminatory system shaped by the white gaze, as programmers and designers encode their judgements into technical systems. What many call algorithmic bias may be more appropriately described as algorithmic violence because of how it brutally targets Black and brown communities.

One of the most dangerous forms of community-level algorithmic violence is “predictive policing,” a range of data-driven surveillance practices that turn entire neighborhoods into vectors of criminal probability. This technology involves the use of software to determine who is considered “criminal” and where crime is “likely” to happen. According to a study of 13 jurisdictions that currently use predictive policing systems, the data on which these systems are built is deeply flawed due to “systemic data manipulation, falsifying police reports, unlawful use of force, planted evidence, and unconstitutional searches.” However, the Stop LAPD Spying Coalition says the problem is not simply dirty data–it’s the fact that predictive policing AI is fundamentally designed to police Black and brown bodies, communities and land.

The coded gaze of predictive policing is an extension of historical systems of racist policing under white supremacy and settler colonialism. This system is not broken, it is working exactly as intended, now with modern technologies to facilitate and automate these processes. Over the last decade in Los Angeles, the LAPD hired Predpol, a predictive policing vendor, to create a statistical model for predicting crime in geographic zones with low-income communities of color, which it labeled “hot spots”. The LAPD combined this with a mapping system, LASER, to create Chronic Offender Bulletins (COBs), which identify people for targeted surveillance. COBs are then analyzed by Palantir, a data-mining search platform that cross-references information from multiple databases and automated license plate readers (ALPRs). Palantir assigns a “score” to persons on the COB list according to gang affiliation, parole or probation status, arrests, and other so-called “quality” police contact. Both Predpol and Palantir operate in other cities where police departments have significant records of racist brutality and misconduct. Contrary to promoting community safety, the Stop LAPD Spying Coalition argues that, “a feedback loop is created where an increasingly disproportionate amount of police resources are allocated to historically hyper-policed communities.”

The criminal legal system’s growing reliance on pre-trial risk assessment algorithms deepens carceral control and racist biases over people’s lives. The Correctional Offender Management Profiling for Alternative Sanctions (COMPAS), an algorithmic system for predicting recidivism rates among pre-trial defendants, has erroneously predicted “high-risk” for Black defendants and “low-risk” for white defendants. Much like electronic monitoring (discussed below), pre-trial risk assessment algorithms are presented as an alternative to traditional detention policies like cash bail. Yet, they threaten to further entrench people’s entanglement in the criminal legal system due to the racial biases and decades of racist criminal history data embedded within these tools. Between the continuou development of surveillance technologies and the expansive, often unmitigated carceral reach of the state, it is paramount to connect historical legacies of criminalization with contemporary inequalities regarding mass incarceration.

Surveillance technology is extending the reach of the carceral state beyond physical prisons by normalizing what MediaJustice (MJ) calls “e-carceration”—the use of wearable electronic monitors, such as ankle monitors, that restrict the freedom of movement and agency of individuals on parole and probation and convince the public that the government should be permitted to track people accused of and convicted of crimes. The sharp rise in the number of people on parole and probation has accompanied mass incarceration; approximately 4.5 million people in the US are under some form of correctional supervision outside of formal jails and prisons. More and more systems, including immigration and juvenile detention, are using electronic monitoring devices.

GPS tracking data from electronic monitors can be incorporated into other agencies’ databases, casting an even wider net of surveillance over a person’s every move. Business is complicit in the explosive development of location surveillance phone apps that may ultimately replace ankle technology. For example, over the last year, BI, the world’s largest electronic monitoring company, has doubled the number of people under ICE supervision with SMARTLink, a social media electronic monitoring app. Governmental agencies and the industries producing electronic monitoring devices have responded to critics of ankle monitoring by going mobile.

While reform advocates tout electronic monitors as an alternative to incarceration, carceral technologies actually widen state surveillance and punitive control over formerly incarcerated people’s lives. They effectively create “digital prisons” that further mass criminalization. The majority of people under e-carceration are confined to their houses and are not allowed to leave without a court order or permission from a parole officer. This type of virtual, often solitary confinement greatly affects the mental and physical health of QT2SBIPOC people. A Black trans woman living in Chicago shared her story of e-carceration with MediaJustice, reporting that while she was under electronic supervision, she was denied permission to leave her house to buy food and fill prescriptions

for HIV medications she requires daily. In emergency situations, people are forced to choose between risking going back to prison for unauthorized movement or, for example, taking a sick child to the hospital.

E-carceration also creates barriers to employment because the restriction on movement prevents people from going to job interviews or working in environments like concrete buildings that may interfere with the monitor’s signal. Electronic monitoring further harms communities already vulnerable to mass incarceration, contributing to financial and other material dependence on the families and social networks of those on parole and probation.

Electronic monitors spread carceral logics into other spaces of society, including the surveillance of workers in factories and of public areas. This expansion of surveillance extends mass criminalization’s reach to low-wage workers. Technology companies that develop surveillance technology already deploy these products on their own workforce. Major companies—like Amazon and Fitbit—have begun using surveillance technology that tracks the motions of its warehouse workers , who are disproportionately Black, migrant, and People of Color in low-wage positions. In outdoor spaces, monitoring devices facilitate “e-gentrification” and create de facto segregation by restricting those who are forced to wear them from entering certain areas. As a researcher and educator interviewed noted, these exclusion zones apply to people who face criminalization in particular ways, such as those on the sex offender registry, who are banned from going within a certain perimeter of public parks. That electronic monitors come programmed with exclusion zones sets a dangerous precedent for other devices that people use daily. This is becoming more apparent with the fear of stingray cell simulators and contact tracing of COVID-19 being used to track down organizers at the 2020 mobilizations in support of Black Lives.

Understanding the hidden ways in which data flows from our personal and private devices to state-based agencies helps us understand the extent to which our communities are being policed and surveilled. The Stop LAPD Spying Coalition defines the “stalker state” as a network of overlapping data-sharing systems between social media corporations, private security firms, public service institutions, military departments, federal agencies, and local law enforcement. It enables data-sharing among state and local police, intelligence agencies, and private companies while also fueling profit-driven tech development as positive progress. A prime example of the stalker state’s reach and impact on a public narrative about safety is the political debate surrounding the militarization of the US border—what so-called progressive politicians espouse as the “smart border”—a surveillance network of AI, drones, cameras, and infrared sensors, as a more humane alternative to the Trump administration’s draconian border wall project.

Tech companies are complicit in developing facial recognition technologies. For example, companies like Thorn and Marnius Analytics have programs that scrape data from escort ads, without consent, to use in the design of new facial recognition technologies targeting sex workers. Under the pretext of “anti-human trafficking” initiatives, companies are enlisting lower-wage workers to surveil sex workers and share data with law enforcement, as one interviewee shared, “this suite of AI tools as well as anti-human trafficking trainings at Uber and Marriott on how to “spot” human trafficking survivors encourage some of the lowest-paid employees of these companies to snitch on sex workers, creating a dynamic where workers are snitching on workers. Sometimes these reports at Marriott and Uber lead to calling of ICE. Once you’re in this system, the state can surveil you using these technologies, harming workers all around.”

US government agencies and local law enforcement are capturing and weaponizing personal and community-level data to increase surveillance and repression of movements. As more data is captured and shared, it is becoming a potent weapon the state wields to disrupt movements. This is happening within the larger international context of US military wars and inter-governmental and inter-agency data sharing in militarized zones such as the US-Mexico border. These practices are then applied by police at the local level in cities across the US. During protests and direct actions, police use surveillance devices like “stingrays” to disrupt activist communications, steal people’s data from their cell phones, and track their physical locations. At least 75 police agencies in 27 states use these devices to capture personally identifying information, which is then shared with other agencies to profile and monitor organizers. These technologies have especially targeted the Black Lives Matter (BLM) movement. During the Eric Garner protests in New York, the NYPD infiltrated BLM and gained access to organizers’ text messages. In 2017, an FBI intelligence assessment surfaced a new designation of security threat, the so-called “Black Identity Extremist” (BIE). In 2020, police used facial recognition technology and the surveillance of social media to identify and arrest activists who were at protests in support of Black Lives. For example, using Clearview AI, activists in New York City and Miami were arrested after participating in the summer protests.

Battle for privacy-first policies and technologies includes protecting access to technologies like end-to-end encryption and protection from digital attacks. Encryption technology is an important safeguard against state surveillance and is under threat. Recent US legislative efforts (i.e.,the EARN IT Act, the LAED Act and the PACT Act), coupled with subsequent defunding of open-source encrypted technologies, show the US government’s transparent desire to access communities’ private communications and do away with access to technological tools employed for privacy like end-to-end encryption (E2E). If signed into law, it will have a global impact on civilian access to encrypted technologies. The loss of this protection would mean that the state would be able to monitor all communications from any device, anywhere. This would imperil the digital security not only of social movements, but of all. The end of E2E would create a backdoor for law enforcement, and also make it easier for civilians with malicious intent (such as abusers, people spreading revenge porn, etc.) to access previously secure communications. The government is also moving toward targeted malware attacks, known as the Network Investigative Technique (NIT) and dragnet malware to capture data from large groups via a single warrant. Additionally, the government is funding tech companies to design malware products that destroy built-in security mechanisms, which millions of people—including organizers—depend upon. QT2SBIPOC organizers report that the state is using these methods to break into their cell phones and personal devices to try to gain access to their data.

Social media intelligence (SOCMINT) companies have found a lucrative niche in contracting with federal and state agencies to spy on organizers. Private firm ZeroFOX monitored the social media accounts and geolocation of BLM organizers during the Freddie Gray protests in Baltimore, referring to organizers as “threat actors” in intelligence reports to law enforcement. Similarly, Geofeedia obtained user data from Facebook, Instagram, and Twitter to monitor activists during the Michael Brown protests in Ferguson. Movement organizations often rely on social media to communicate with members and promote their campaigns and events, which makes them vulnerable to surveillance, infiltration, and data poaching. They acknowledge that a major contradiction they face in using mainstream platforms is what some call the “network effect.” That is, as organizers reach more and more people on these platforms, they also become more dependent on corporate infrastructure for their organizing work. This, in turn, reinforces the status of mainstream platforms as the “default” venue for all communications. Organizers interviewed for this research even report that police and federal agents have used fake social media accounts to pose as activists and try to gain access to personal and organizational information. The ways that private companies collaborate with state actors is often opaque, but the consequences of these collaborations are very real.

Third-party agreements between government agencies and data-brokering companies facilitate the interlinking of digital and physical threats. Border agents now commonly ask people for their social media accounts upon entering the US and there are reports of sex workers being denied access at border crossings because of social media and escort ads. Migrant justice advocates warn that immigration authorities monitor queer and trans asylum seekers’ posts on platforms like Facebook and use their social media content to argue against their asylum cases. The linkage of digital and physical surveillance through third-party data sharing has enabled ICE to conduct targeted raids based on people’s addresses, social media posts, and location data. In response to concerns about ICE raids in Puerto Rico, some organizers reported having to close their social media accounts, avoid posting about their activism, or cover their faces during protests to avoid being deported.

The increasing collaboration between the US government and private tech companies is fueling a system where corporations are profiting from surveillance products. This is a clear example of how ‘surveillance capitalism’ works in the service of criminalization. The Department of Homeland Security (DHS), CBP, and ICE are spending billions of taxpayer dollars annually on contracts with tech companies to target immigrants of color. Nowhere is this more apparent than with the escalating arrests, detentions, and deportations that comprise the US administration’s ongoing war against migrant communities. This war, which also includes the geographical and biometric surveillance of immigrants and their communities, represents what has been called crimmigration: the intersections of criminalization and immigration. As with other forms of criminalization, the use of technology to consolidate power and capital is well documented within the policing and militarization of the US borders and immigration system.

The growing technical interconnectedness between federal agencies and local police departments is eroding protections granted by “sanctuary cities” that have historically limited the use of federal immigration detainers for a person of undocumented status in custody. Due to inter-agency data sharing and police use of social media, ICE agents are surveilling and arresting migrants in churches, courtrooms, hospitals, and other public spaces with greater frequency. This restricts the movement of people of undocumented status and prevents them from accessing much needed health, legal, and social services for fear of being detained and deported.