For decades, parole and probation officers have supervised people on house arrest, parole, and probation remotely with GPS-enabled ankle bracelets. Now, as a pandemic deters person-to-person contact, remote supervision in criminal justice is moving to a new frontier: the smartphone.
Services such as Tracktech mimic more well known forms of remote supervision like ankle monitoring, but instead are installed on phones as a condition of release. These apps monitor juveniles and adults alike on parole, probation, and pretrial release. Parolees can connect with case officers through text or video. GPS features track whether a parolee has kept a consistent schedule, moving only between work and home. Officers can place geofences around “off limit” spaces, like liquor stores or gun stores, and receive notifications if a parolee gets too close.
These startups claim remote monitoring is safer amid the pandemic, and helps people avoid violating parole conditions. But critics call the field a form of “e-carceration,” a sanitized, but still dehumanizing form of imprisonment. As more corrections departments consider the tech, a group of researchers has begun piloting a predictive, artificial intelligence-enabled version of remote supervision that already has ethicists raising alarms.
“One of the things we need to think really hard about is the [surveillance] aspect of electronic monitoring,” says Chaz Arnett, a law professor at University of Maryland researching technology and decarceration. He argues that the prevalence of remote monitoring is in part an effort to diffuse responsibility in the criminal justice system.
“No judge wants to be that person who released someone after sentencing and then something happens,” he says. Electronic monitoring “gives them some level of cover. Judges can say ‘Yes, I did release this person, but I placed them on electronic monitoring.’ So whoever is in charge of electronic monitoring is at fault for whatever happens.”
That would usually be probation or parole officers, who must monitor a lengthy and shifting list of conditions for parolees: meeting with sponsors and supervisors, surprise drug tests, proof of a steady job and stable housing, avoiding liquor stores and attending support groups, among others.
These conditions are ostensibly meant to keep recently released people on the straight and narrow, but in practice they become their own set of responsibilities. Parolees risk a “technical violation” if they run afoul of these rules, a one-way ticket into jail or prison. Supervisory apps make it easier to explain extenuating circumstances to an agent—you need to work overtime to pay rent, but that means missing an AA meeting, for example—but also increases surveillance, Arnett says. “You end up being in greater contact with them than you would if you weren’t on this level of monitoring,” he explains.
These technical violations drive huge numbers of people into jails and prisons. Nationwide, almost one in four admissions to state facilities are due to technical violations, according to the Council of State Governments.
This is the first problem these apps say they’re trying to fix: Automating reminders and sending notifications to reduce the chance that parolee misses a meeting or check-in with a sponsor, supervisor, or job recruiter, and is sent back to prison. They also have messaging systems so users can instantly reach supervisors.
Jacob Sills, CEO of Uptrust, another supervision app, says he set out to create a simple reminder system for public defenders and their clients after learning that a single missed meeting could drastically disrupt someone’s life.
Uptrust eventually pivoted to working with parole and probation officers, creating an app that allows for video chat, automated reminders, and direct messaging between parolees and officers. Uptrust does not track its users’ location.
Essentially, the platform has two functions: communication between parolees and officers and notifications to remind parolees of upcoming appointments and other parole conditions. Sills says he saw other apps that “were just building tools to extend mass incarceration from cages to people's mobile devices,” but didn’t seem to care about a parolee’s well-being.
That is the second problem these apps are trying to fix: officers are overloaded. Statistics are hard to come by because probation and parole are handled by states and counties. The American Probation and Parole Association recommends a ratio of 200 low-risk parolees to one officer. Officers have complained of overload, however. In Mississippi, officers claimed they faced caseloads of around 300 probationers and parolees. A bill to limit the number to 100 failed this year. Arnett argues that the system is built to make recidivism the fault of officers and parolees.
Tracktech, another vendor, pairs its messaging and automation software with more enforcement capabilities (location tracking and geo-fencing alerts, for example) and a case management dashboard for supervisors. It is essentially a ranking system. If an officer has 100 parolees to supervise, for example, the apps can offer an “early warning” system of cases that need attention.
“It's sort of a red-yellow-green triage based on what's happened in the last reporting period,” explains Tracktech CEO, Michael Hirschman. Tracktech integrates with parole systems to rank and flag parolees. A parolee may be labeled red, or high risk, for going near an ex who's filed a restraining order. Yellow would be more for someone missing or late to an appointment with a reasonable excuse.
But this leads to the dilemma that concerns Arnett, where constant tracking creates the need to constantly ask permission from supervisors. James Kilgore, an activist and author who spent 6 ½ years in prison, says this need to constantly update and remain in touch with a supervisor is invasive. The pressure to remain accessible makes it harder to form any real connection with the officer tasked with helping them re-enter society.
“The criminal legal system is looking for cheap ways to do their job instead of recognizing, first of all, that punitive supervision does not work,” he says. “It makes people more likely to hide their real feelings, knowing that they might be punished for it.”
Kilgore leads a re-entry program in Illinois, and says money spent on remote supervision is a misuse of funds. Instead, he suggests states provide people with social workers, therapy, and access to housing.
“I'm dealing with individuals on a daily basis who are navigating the hurdles to reintegrate into the local community,” he begins. “These people do not need technology. They don't need tracking. They don't need reminders. They need human beings and access to resources. And those are all more expensive than cell phone apps.”
Noting the lack of resources to fund robust, person-to-person support, one team of researchers is exploring an approach combining technology and cognitive behavioral therapy. Last year, the National Institute of Justice granted $1.9 million to the researchers, from Florida State University, Purdue, and the University of Alabama.
The group wants to create what Carrie Pettus-Davis, an FSU professor and one of the project’s leads, calls a “game of life app,” that combines remote monitoring, artificial intelligence, and biometric data from users.
“One of the really critical intervention techniques for people in recovery is to help them tune in to the biological factors that lead to relapse,” Pettus-Davis says.
The researchers plan an app that will measure a person’s heart rate and breathing rate. When those increase, that can signal stress, such as a temptation to use drugs or alcohol. The user may not know their own triggers and the app could, in theory, notice them and help build better habits.
“What substance abuse researchers have found out is that you can prompt using applications,” she says. “You can prompt them, then send the message, ‘You’re triggered, call mom.’” She cites a client who struggled with an addiction to crack cocaine. During support therapy, the man, a heavy smoker, mentioned that the sound of his lighter clicking often triggered cravings for the drug. By focusing on his smoking habit, he made strides in curbing his addiction.
The system might suggest different actions for different stresses: Meditation apps for agitation at work, for example, or reminders to call an AA sponsor if they struggle with the urge to drink. Pettus-Davis says that the app might also change the thinking of parole and probation officers.
“Officers don't know much about behavioral health disorders,” she says. “They misinterpret the evidence of a health disorder as indicators of future engagement and criminal activity, and they over-respond,” which can lead to some people being returned to prison.
To prevent people from being further criminalized, researchers say they plan for only the parolee or a therapist to see any biometric data. Umit Karabiyik of Purdue University, another researcher on the project says the system aims to help parole officers “reduce the burden to handle these participants.”
Kilgore thinks the researchers are chasing another shiny tech object, instead of providing the personal support that has proved effective. “In order to figure out if somebody is likely to do something self-destructive or destructive to others, build a human relationship with them,” he says. “You build relationships so you can tap into their feelings and so they trust you and share what their problems are.”
Updated, 11-13-20, 9pm ET: This story has been updated to clarify how Uptrust works.