“Mail digitization programs impose a heartbreaking burden on incarcerated people and their loved ones,” said Stephanie Krent, a staff attorney with the Knight First Amendment Institute. “While correctional facilities have an obligation to make prisons safe, those concerns should not be used as an excuse to subject these communities to ever-more invasive surveillance.”

In March of 2020, the BOP launched a pilot program that digitizes incoming mail. Under that program, a private telecommunications contractor, Smart Communications, opens and scans all non-legal mail sent to at least two federal correctional facilities. The company destroys all the original letters, pictures, and photos that are sent, but retains electronic copies, along with additional information about the senders, in a searchable database. The pilot program was supposed to last a year, but the agency has not publicly disclosed the current status of the program. It also has not explained whether it has policies in place to protect the data collected from misuse or to ensure the data is not retained in Smart Communications’ database indefinitely.

The lawsuit notes that while Smart Communications markets its service as a method of improving safety and security in correctional facilities by preventing contraband from entering a jail or prison, data collected after the Pennsylvania Department of Corrections adopted the Smart Communications system showed a slight increase in drug-test positivity rates, indicating that the mail was not the only or even main source of drugs entering the facility.  

“The Bureau of Prisons must provide a full accounting of this new form of surveillance,” said Xiangnong Wang, a legal fellow with the Knight First Amendment Institute. “The public has a right to know the policies governing the bureau’s review, retention, and dissemination of such deeply personal information, and it has a right to know whether the program has actually succeeded in any of its goals.”

The lawsuit asks the court to require the BOP to provide information about when it began this program, whether it plans to extend it, and what it is doing with the data it collects.

Read today’s complaint here.

Lawyers on the case include, in addition to Krent and Wang, Alex Abdo of the Knight First Amendment Institute.

For more information, contact: Lorraine Kenny, communications director, lorraine.kenny@knightcolumbia.org. 

How do mail digitization and surveillance programs, or for that matter phone surveillance, work in the U.S. penal system? 

Stephanie Krent

With mail digitization, we’re talking about the process of taking an original, physical piece of mail—a letter, a drawing, a greeting card—and creating a lasting digital copy the way you would if you were scanning your own files. But in this instance, it’s digitization that’s being conducted by the prison facility itself, transforming the physical piece of mail into a text-searchable file. Generally, the original mail itself is destroyed and the incarcerated recipient gets a photocopy or a scanned image of the mail rather than the original piece of mail itself.

Phone surveillance is similar, in that it is transforming an audio conversation into a text-searchable digital file. Typically, a facility or a contractor records phone calls as they occur throughout the day and uses software that creates an automated digital transcription, often using natural language processing, that will create a long-lasting digital file memorializing that conversation. 

In general, the constitutional authority of a prison official to listen in on a phone conversation or to read a piece of mail has been well-established. The difference here is that the creation of a digital file, which can live on indefinitely on a facility’s servers, opens up a whole host of new surveillance capabilities. Instead of glancing at a letter in real time, officials could come back to a letter to read it years after it’s sent. They can use keywords to search preexisting files, and in some instances to request a notification any time that a keyword is used in a particular conversation or letter on a forward-looking basis. With stored data on senders and recipients, they can map out networks of communications based on who is writing letters or speaking to whom in those facilities. Essentially, while prisons have a responsibility to keep the people in their facilities safe, mail digitization creates a vast trove of new information that can be used for any purpose within the facility. 

What’s the rationale for using these new technologies or surveillance? 

It varies depending on the facility itself and the type of technology that’s being introduced. For mail digitization, for example, the most common rationale is a fear of contraband being introduced into the facility, for instance by soaking a synthetic drug onto an envelope or a stamp. Facilities have tried to eliminate that risk by denying original mail to people who are incarcerated.

Once a facility has the capability to conduct this surveillance, the mission tends to creep, with officials coming up with more and more supposed benefits to it.  

These programs are often announced based on very narrow or particular goals. But once a facility has the capability to conduct this surveillance, the mission tends to creep, with officials coming up with more and more supposed benefits to it, including vague invocations of efficiency. The technology becomes part of the fabric of an institution and can be used in ways that are completely untethered from the original, narrower justification.

Are these technologies already widely in place?

Yes, they are available in many different states and are growing rapidly. There are several local facilities that have been digitizing mail, and, in the last few years, statewide penal systems have also started to adopt these technologies. Pennsylvania was the first state to adopt a mail digitization system through the contractor Smart Communications in 2018. The Federal Bureau of Prisons also began a pilot program with Smart Communications in 2020 that is the target of our FOIA request. In the last year and a half since that pilot began, we’ve seen an increasing number of states move towards adopting mail digitization. North Carolina began digitizing mail. Wisconsin this fall also announced a program to use the same contractor North Carolina uses, Text Behind, for their incoming mail system. Florida and Massachusetts have both had hearings on whether to adopt mail digitization. What we’re seeing is these programs gaining a foothold and becoming more emergent in recent years.

Why has the Institute focused its attention on the federal Bureau of Prisons pilot?

There are two reasons we wanted to target our FOIA request on the federal pilot program with Smart Communications. First, at the time we filed the request this spring, the pilot was actually supposed to have ended. But when we went on the bureau’s website, it appeared it was still ongoing. Yet there was no announcement from the bureau that the pilot was going to be extended, or for how long. There was no real information about what it was doing and how it was using the mail that was part of this pilot program. So we felt that there was a real inflection point, before the program was fully entrenched and adopted within the bureau, when we could help the public understand exactly what the mail digitization at these prisons entailed. 

The second reason was that given the bureau’s size and status, its policies can be very influential nationwide. Its decision to digitize mail wouldn’t affect only the more than 150,000 people who are in bureau custody, but would affect the nearly 2 million people who are incarcerated nationwide. By learning more about the bureau program and by striking at this moment, the information we learn might influence the way these technologies are being considered across the country.

What are the primary concerns the Institute has about the prison mail surveillance program? 

There are at least two facets of mail digitization that are disturbing. One is the loss of access to original mail. You could have a system that denies incarcerated people original mail, but that doesn’t incorporate enhanced surveillance or digitization. In fact, some states and carceral facilities have adopted models that involve photocopying without scanning and long-term storage. Even that model has serious implications for expression behind bars, because photocopying can delay the time it takes mail to get delivered. And if copying is done haphazardly, it can make mail and photographs unintelligible. That damages the quality of communication and can weaken individuals’ ties to their community. Those concerns are present in both photocopying and mail digitization regimes. 

Another disturbing aspect of mail digitization is the surveillance effect. Facilities sweep up all of the incoming mail for every incarcerated individual in the facility and retain it indefinitely. We don’t know exactly how those records are being used, and it likely varies from institution to institution, but we have yet to see meaningful limits placed on how prison officials could use this information. What we do know is that the contractors who routinely provide these services explicitly advertise them as surveillance tools, with the ability to store mail long-term, to help facilities come up with search terms that they can use to track and target correspondence, and to create “profiles” on senders and recipients of mail. 

Letter-writing, especially to someone incarcerated, can be very personal, very intimate. Mail digitization ... has a dramatic chilling effect on people who want to communicate with someone behind bars.   

Letter-writing, especially to someone incarcerated, can be very personal, very intimate. Mail digitization ratchets up the personal cost of that communication. If you want to send a letter into one of these facilities, you have to open up your thoughts not only to the intended recipient but also to anyone else who is working at the facility, who is a contractor at the facility, or maybe who works at a law enforcement organization that has a data-sharing agreement with that facility. All of those people may be reading your letter as well. They may be able to read or reread or search it years later for any reason. That has a dramatic chilling effect on people who want to communicate with someone behind bars. 

So one concern is that it’s not just the prison officials or the vendor reviewing these digitized mail records, but others at outside law enforcement agencies too?

That’s right. There’s not enough that we know yet about how prison officials are actually using retained mail and how they’re sharing that information. Records released to journalists at Reuters recently revealed that Suffolk County, New York, shared information gleaned from its phone transcription and surveillance system with more than two dozen entities. 

We don’t know yet whether there was some sort of analogous data-sharing agreement in the federal pilot program, or in other states that have adopted similar technologies. That’s something to watch out for because of the concern that, despite public justifications relating to concrete issues within the prison facility, these tools will actually be used as general law enforcement tools, mining sensitive communications for whatever reason a law enforcement officer might think of. When you think about the creep of the surveillance program and the chilling effect that that creates, that’s quite dangerous. 

Are family members, even children, themselves being surveilled?

Yes, and that’s by design. Often, children send letters, cards, and poems that are being scanned and stored. These children and other family members are caught up in long-term carceral surveillance merely because they love someone who is incarcerated, and have sometimes even stopped sending mail as a result.

Certain vendors have actually advertised enhanced surveillance capabilities with respect to senders of mail. As we explained in our FOIA request, Smart Communications has what it calls a Smart Tracker tool that it says creates profiles on and stores details about the senders of mail. For example, if senders log in to the Smart Communications website to see if their letter has been delivered, Smart Communications will log the GPS location and the IP address where that connection occurred and then store it. They’ll also store any email addresses used to track mail. And they’ll store any associated phone numbers or addresses they can find for the sender. Then they’ll create a running list of every incarcerated person the sender has communicated with. 

Essentially, it's creating complete surveillance profiles not only of people who are incarcerated but also of people who are not incarcerated, who just want to communicate with those who are behind bars.  

Essentially, it’s creating complete surveillance profiles not only of people who are incarcerated but also of people who are not incarcerated, who just want to communicate with those who are behind bars. This surveillance system spreads far beyond the prison walls. And what’s really scary is that, in the Smart Communications proposals I’ve reviewed, there is no mention of a carve-out to protect children from this kind of tracking.

What do we know about the retention policies for these records? 

We know very little about how facilities are retaining these records in practice. Frankly, it wouldn’t surprise me if many of these facilities have no policy on this at all and are just trying to figure it out as they go along. But two of the main contractors who advertise mail digitization systems, Smart Communications and Text Behind, advertise to facilities that they will keep mail for at least seven years. And that includes the time after someone who’s incarcerated has been released, even if they were just incarcerated for a short period of time. 

It’s also possible that the retention is far greater than what these vendors are actually advertising. In 2018, the Smart Communications CEO was asked about this and said that in over 10 years, Smart Communications had never deleted a single record, and that they were all available for investigators. So, in effect, essentially once a facility is able to lay eyes on something, it is able to keep that and use it in any way forever. That’s frightening. It’s extremely concerning that these facilities are given these surveillance capabilities yet have not announced safeguards to ensure that this isn’t just another dragnet scheme. 

But if there are legitimate safety concerns, how do you balance this with concerns about free expression? 

From a constitutional perspective, there’s a balancing test. There is significant discretion given to correctional facilities to enact rules to protect safety and security, but those restrictions really have to be rationally related to legitimate penological goals. 

Many facilities have pointed to the introduction of drugs through the mail as the reason these programs have been put in place. But in certain instances, mail digitization systems have not meaningfully moved the needle on contraband use in prisons and jails. The American Prospect reported this year that in Pennsylvania, the adoption of mail digitization didn’t lead to a concomitant decrease in the drug test positivity rate in statewide facilities. So there may be reason to be skeptical of these claims and to demand that if a facility is digitizing mail, it should justify that by pointing to data that can be analyzed by outsiders and by advocates, to ensure that these programs really are tethered to legitimate goals. And facilities should do so not only when the system is first introduced but at regular intervals, so that advocates can make sure that these systems aren’t being abused.

What is the status of the Institute’s FOIA lawsuit seeking records on the Bureau of Prisons’ pilot? What is the Institute trying to learn?

We filed a FOIA request in June 2021 for records relating to the bureau’s pilot program, and any other instances of mail digitization that facilities within the bureau umbrella have embarked on. Our goal was to understand how these programs work in practice, because vendors are promising the moon and giving facilities significant new surveillance capabilities. 

There’s a great deal we don’t know. We don’t know how facilities within the bureau are actually using the surveillance capabilities at their fingertips. We don’t know how long records are retained. We don’t know whether or not there’s a certain level of suspicion that’s required before records are searched. We don’t know how records are being shared across facilities or with law enforcement who are not part of the Bureau of Prisons. 

We also don’t know whether there are any safeguards in place to prevent abuses of the system or to prevent self-serving activities. For example, we’ve learned of incidents when surveillance technologies were being used to study phone conversations about the facility’s COVID protocols in an attempt to shore up evidence that could protect the prison from any future complaints that it was not doing enough to prevent the spread of COVID behind bars. That anecdote suggests that, at least in those facilities, there aren’t really any limits on how officials can use this technology. 

We had all of these open questions that would impact the way we viewed the pilot program’s impact on free expression. And because the bureau hadn’t answered these questions publicly, we felt the need to file a FOIA request. Then when the government didn’t respond to our FOIA requests within the statutory deadline for responding, we filed a lawsuit. Now, through negotiations with the Bureau of Prisons, we’ve received our first two sets of records. We’re expecting at least one more. There’s still a lot left to be learned about these programs, which we’re hoping to uncover in the next few months.

Visit our case page for the latest updates and more on Knight Institute v. Federal Bureau of Prisons.

Research for this report was contributed by Hye Jin “Jinny” Lee, legal extern at Knight Institute.

In at least one TextBehind contract, the company pledged that it would retain scanned copies of all mail for at least seven years, including both incoming and certain outgoing mail, and even after the letter’s recipient has been released.

This type of prison mail surveillance is growing quickly across the United States. I’ve written before about the significant First Amendment concerns these programs raise, including the chilling effects of pervasive surveillance, the burden on expression behind bars when mail is scanned incorrectly, delayed, or lost, and the irreplaceable nature of physical mail for those who have limited connections to the outside world. You can read that analysis here.

But in recent years, prison mail has come under threat. Many correctional facilities have begun digitizing physical mail, denying incarcerated people the original mail sent to them, and subjecting both the senders and recipients of those letters to heightened surveillance. These programs have found footing in both local jails and statewide correctional agencies; even the federal Bureau of Prisons began a pilot program with Smart Communications, one of the primary telecommunications contractors offering mail digitization services.

Earlier this summer, the Knight Institute filed a Freedom of Information Act (FOIA) request with the Bureau of Prisons to understand how federal corrections officers are storing and using the trove of personal information collected by Smart Communications. Because the bureau didn’t respond to our request, we filed suit last month.

Since then, we’ve learned that the bureau has paused its pilot program, although it is still actively considering expanding the program in the future, with the support of BOP union heads. We’re litigating over this request to ensure that before the bureau makes a final decision, it has provided the public with a full accounting of the pilot program, including the data it collected, the policies that governed the use of that data, and the safeguards, if any, it put in place to prevent abuse of the system. These records are particularly important given the serious First Amendment concerns presented by this new form of carceral surveillance.

Considering the immense harms suffered when correctional facilities destroy original mail and ratchet up surveillance, the government must do far more to justify their program before resuming mail scanning at any facility.   

Mail digitization systems collect and retain deeply personal information, not only about those in prison, but also about their loved ones, their educators, and their religious advisors. Facilities that have contracted with Smart Communications, for example, direct letter writers to send all incoming non-legal mail to a site in Florida, where Smart Communications employees scan and upload images of the mail into a searchable database accessible to employees at correctional facilities. Correctional officers review the incoming mail in the database and forward a digital copy or photocopy of approved mail to the recipient. Originals, including handwritten letters, drawings, and photographs, are never shared.

Through its “Smart Tracker” service, Smart Communications also surveils the senders of mail who attempt to track their letters, storing information about senders’ IP addresses and GPS locations, their contact information, copies of letters they’ve sent, and a list of all “connections”—that is, every incarcerated individual the sender has communicated with. Long after the copied mail is delivered, these electronic records live on in Smart Communications’ database, never more than a click away from corrections officers. The company’s CEO told a reporter in 2018 that “Smart Communications has never lost or deleted any records or any data from our database. There are hundreds of millions of data records stored for investigators at anytime.”

Because the bureau has not turned over any documents in response to our FOIA request yet, we don’t know how exactly its pilot program worked. But it’s clear that the pervasive surveillance enabled by Smart Communications and vendors like it has already chilled the expression of members of the public who wish to send mail to those held in correctional facilities. In response to Pennsylvania’s adoption of the Smart Communications system in 2018, for example, the spouse of one incarcerated individual said that she would not “send any pictures to [her husband] that have faces of our children, our grandchildren,” because she and her husband “want … no family faces saved in that database whatsoever.” Another spouse stopped sending her child’s original artwork to her husband serving a life sentence.

Mail digitization systems have also erected unnecessary barriers to meaningful communication, further isolating people in prisons and jails. People in facilities that have adopted or tested such systems report long delays in receiving mail and incorrect copying, including distorted photographs, illegible letters, and lost pages. And the loss of original mail weighs heavily on them. Gregory Marcinski, who has been incarcerated for almost 20 years, told a reporter last year that he could still smell the scent of his father’s favorite cigarettes on the last letter he received before his father passed away in 2016. He treasured the ability to “see the ink he used on that paper,” especially as he could no longer receive original letters from his partner or her children.

These programs may also violate the confidentiality standards required for certain sensitive communications. Just Detention International, a human rights organization, has argued that these systems are inconsistent with the national standards set under the Prison Rape Elimination Act requiring correctional facilities to “enable reasonable communication between inmates and [victim advocacy] organizations and agencies, in as confidential a manner as possible.”

To be sure, correctional facilities enjoy significant leeway when they create programs that are designed to keep prisons—and the people in them—safe. But this discretion is not limitless: Restrictions on the speech of those incarcerated must be rationally related to the government’s goals, and must leave alternate means of communication open. Unfortunately, there is little evidence that these invasive programs have any meaningful impact on security. Although ostensibly designed to curb the introduction of contraband into facilities, the adoption of mail digitization in Pennsylvania did not lead to a decrease in the average drug test positivity rate. And reporting suggests that correctional officers—not letter writers or even in-person visitors—are primarily responsible for the introduction of contraband to correctional facilities. Considering the immense harms suffered when correctional facilities destroy original mail and ratchet up surveillance, the government must do far more to justify their program before resuming mail scanning at any facility.