The Case Against Using Artificial Intelligence to Monitor Prisoner Communications

Carly Cato*


* J.D. Candidate, 2026, Indiana University Robert H. McKinney School of Law— Indianapolis, Indiana; B.S., 2023, University of Southern Indiana—Evansville, Indiana. Thank you to my faculty advisor, Stevie Pactor, for her guidance and advisement on this note.

Introduction

Monitoring prisoner telephone calls is considered an essential function for prisons across the United States, as calls represent a main form of communication between prisoners and the outside world.[1] While most prisoners use their telephone privileges for lawful purposes, some use it as an opportunity to facilitate criminal activity and traffic drugs or weapons into prison.[2] As a result, prison systems across the United States monitor inmate phone calls as one of many measures taken to ensure safety and security within prisons.[3] However, because of understaffing issues in prisons across the nation, inmate phone calls often go unmonitored.[4]

New technology with advanced data collection capabilities attempts to solve this issue. Several states have employed AI-based technology to automatically monitor prisoner phone calls.[5] The technology, using “keyword-based searches,” monitors inmate phone calls and alerts administrators when an inmate uses a specific word or phrase.[6] Although AI has been moderately successful in assisting facilities in the states that have already implemented the technology, there are concerns that its capabilities will lead to further erosion of prisoners’ rights.[7] Technology may be able to improve prison security, but not without a cost to prison inmates. Awareness that the government is watching or listening “lead[s] people to modify their otherwise ‘normal’ behaviors, whether consciously or subconsciously.”[8] This is true even for those individuals subject by law to the highest degree of government surveillance: prisoners.

This Note argues that prisons should not implement AI call monitoring technology because of the potential harm to inmates’ constitutional rights associated with the technology. Part I discusses the judicial limits on the monitoring of prisoner phone calls, demonstrating the wide latitude prisons have in performing this function. Part II reviews other states’ movement towards using AI-based technology for this purpose and describes how the technology works in practice. Part III explores the First Amendment issues raised by using AI for call monitoring. Part IV examines the technology’s effect on prisoners’ substantive due process rights. Finally, Part V argues that the use of AI for call monitoring is not justified under the standard set by the Supreme Court for prison practices.  

I. Limits on Monitoring Prisoner Phone Calls

    “Prison walls do not form a barrier separating prison inmates from the protections of the Constitution.”[9] However, the Supreme Court has made clear that whether an inmate retains a particular right while incarcerated depends on balancing the importance of that right with “legitimate penological interests.”[10] In Turner v. Safley, the Court decided the constitutionality of two prison regulations: one related to an inmate’s marriage rights and the other related to inmate correspondence.[11] The Court in Turner announced a new rule to help determine the constitutionality of prison regulations: “[W]hen a prison regulation impinges on inmates’ constitutional rights, the regulation is valid if it is reasonably related to legitimate penological interests.”[12]

    Relevant to this analysis are four factors laid out in the Turner opinion. The first factor requires “a ‘valid, rational connection’ between the prison regulation and the legitimate governmental interest put forth to justify it.”[13] The second factor involves a consideration of alternative means available for inmates to exercise the asserted right.[14] Third, “the impact accommodation of the asserted constitutional right will have on guards and other inmates” must be assessed.[15] Fourth, whether there are “ready alternatives” for the prison regulation assists in determining whether a regulation is reasonable.[16] In its opinion, the Supreme Court emphasized that prison administrators are owed deference “in gauging the validity of the regulation,” creating a high burden for prisoners challenging the constitutionality of regulations.[17]

    Concerning inmate telephone calls specifically, the Federal Bureau of Prisons has acknowledged that “there is no legal prohibition against restricting inmates’ telephone privileges.”[18] Some federal courts have held that inmates have a minimum right to telephone access under the First Amendment, while others have taken a more limited view.[19] For example, in Indiana prisons, inmates’ telephone privileges are vulnerable to pervasive restrictions by prison administrators, as the Seventh Circuit has ruled that telephone access is not a right guaranteed by the First Amendment.[20] Thus, while Indiana prisoners are allowed to use the telephone while incarcerated, that privilege can be restricted based on legitimate penological interests of the prison facility.[21] This is consistent with the view expressed by the Bureau of Prisons, controlled by the Turner standard, that “restrictions on telephone privileges will not violate the inmate’s First Amendment rights if they are reasonable.”[22]

    II. The Movement Towards Using AI in Corrections

      Although AI was invented several decades ago, recent technological developments have “pushed the boundaries of the way computer machines used to operate and function[] to make human lives easier.”[23] Given its undeniable benefits in streamlining operations across industries, AI will likely transform the corrections industry in the coming years.[24] A brief prepared by a division of the Department of Justice (DOJ) recognized that “AI-enabled tools have the potential to improve efficiency, reduce costs, and expand capabilities across many criminal justice use cases.”[25] At the same time, the DOJ acknowledged there are operational limits to using AI in corrections, and the use of such technology is due careful consideration before it is implemented.[26] Therefore, in assessing states’ movement towards using AI in corrections, it is imperative to balance both the benefits and drawbacks of the technology in light of the potential constitutional harms it poses to an already-vulnerable population of citizens.

      A. How Are States Using AI in Corrections?

        As it relates to the administration of prison facilities, “[u]nmonitored inmate communication ranked third in the list of most pressing priorities in a National Institute of Justice study of the nation’s nearly 7,000 correctional institutions.”[27] This is because some prisoners use their phone privileges for illegal purposes, like facilitating the trafficking of contraband into a facility.[28] At the same time, staffing issues plague correctional institutions across the United States as facilities struggle to recruit and retain staff.[29] As a result of these staffing shortages, prisons are unable to monitor a substantial number of prisoner phone calls, meaning calls sometimes go unmonitored.[30]

        In response to this issue, some states are using AI-based technology to automatically monitor and record inmate phone calls.[31] For example, in New York, the use of AI successfully flagged an inmate’s telephone call where he threatened to kill the judge and prosecutor in his case upon release from prison.[32] In Alabama, AI technology assisted local law enforcement officers in solving a cold case when it flagged phone conversations of inmates discussing committing an unsolved murder.[33] And in Georgia, the Department of Corrections monitored over 7.5 million inmate calls and identified evidence of criminal activity in over 1,600 calls.[34] These outcomes in states where AI is used for call monitoring reveal the possible benefits of using this type of technology in corrections.

        B. How Does the Technology Work?

        To understand the potential issues raised by AI call monitoring technology, it is useful to briefly consider how the technology works in practice. Several of the states that are using AI technology for call monitoring employ“Verus,” a program intended to help law enforcement officers gather intelligence by monitoring and transcribing prisoner phone calls.[35] Verus provides “near real-time transcriptions, with keyword-based searches and alerts, enabled by artificial intelligence.”[36] The technology is based on a natural language processor that focuses on what words are being said, as well as the speaker’s frequency, volume, and tone.[37] Facilities that use Verus “may only search for calls by objective data, such as inmate name, date of call, or keyword”—searches by race, age, gender, ethnicity, religion, or sexuality are not allowed.[38]

        C. Issues Presented by the Use of Such Technology in Corrections

        In light of this movement towards using AI in corrections, this Note examines how Verus and similar programs affect the limited rights retained by prisoners. To demonstrate the possible effects that advanced call monitoring technology could have on inmates’ rights, the Mosaic Theory of the Fourth Amendment, which addresses the issue of increased government surveillance in the age of new technology, is a useful metaphor to illustrate the issues associated with technology capable of mass data recording and collection.[39] While the theory was first raised in the Fourth Amendment context, the same general principle applies to individual privacy outside of the Fourth Amendment: “New technology may provide increased convenience or security at the expense of privacy.”[40]

        This theory, confronted by Justices Alito’s and Sotomayor’s concurring opinions in United States v. Jones,contemplateshow government surveillance, over time, can reveal a substantial amount of personal information based on individually collected data points.[41] The issue before the Court in Jones was whether the placement and use of a GPS tracking device on a crime suspect’s vehicle constituted a “search” under the Fourth Amendment, which the majority ultimately determined it did.[42] The majority’s opinion did not deal with the issue of mass data collection possible with the use of technology, like a GPS monitoring device, but two concurring opinions express an awareness of the possible implications on individual privacy.[43]

        In his concurrence, joined by Justices Ginsburg, Breyer, and Kagan, Justice Alito stressed that new technology has made it possible for law enforcement to remotely collect data that would have previously required large teams of officers and extensive resources.[44] For instance, before modern technology was widely available, surveillance of a crime suspect’s location over twenty-eight days would have required considerable law enforcement resources to gather a similar number of data points; now, GPS trackers, which can be discreetly placed on vehicles, can provide law enforcement with thousands of data points automatically.[45] Justice Alito highlighted that privacy concerns may arise where “dramatic technological change” makes “long-term monitoring relatively easy and cheap,” suggesting that this kind of data collection provides the government with more information than should be constitutionally permissible.[46]

        Justice Sotomayor’s concurrence also touched on this theory of a “mosaic” collection of data. Because modern surveillance technology is more cost-efficient and consumes fewer law enforcement resources, she argues that “GPS monitoring—by making available at a relatively low cost such a substantial quantum of intimate information about any person whom the government . . . chooses to track—may ‘alter the relationship between citizen and government in a way that is inimical to democratic society.’”[47] Justice Sotomayor further argued that “the government’s unrestrained power to assemble data that reveal private aspects of identity is susceptible to abuse.”[48]

        Thus, the concurring opinions in Jones express an awareness of the effect that improving technology can have on government surveillance and individual privacy interests. By acknowledging that advanced technology might not fit squarely in Fourth Amendment search jurisprudence, several Supreme Court justices advance that modern technology can, and potentially should, change the way courts handle individual privacy rights, specifically under the Fourth Amendment.

        In short, the Mosaic Theory “looks at the government’s investigative action and asks whether the type and amount of information gathered, when viewed in the aggregate, is so revealing that . . . even if the individual data points do not reveal that much in isolation,” the government obtains substantial information about an individual over time.[49] “Put in terms of the mosaic metaphor, the individual data point is a singular tile, which viewed by itself is largely meaningless. But when combined with other data points a clearer picture emerges, just like when many tiles are combined to create a beautiful mosaic.”[50] The Fourth Amendment itself is generally not applicable to prisoners given the nature of prisons,[51] but the Mosaic Theory of the Fourth Amendment provides a useful lens for addressing other possible constitutional issues associated with the use of AI for prisoner call monitoring.

        Turning to how this theory plays out with AI call monitoring, courts have held that prisoners have “no reasonable expectation of privacy” in their cells[52] or telephone calls,[53] meaning prison facilities currently have a wide latitude in gathering prisoner information by performing cell searches and monitoring telephone calls. Yet, because these methods of data collection currently rely on the availability of staff to carry them out, the ability of prisons to gather data is somewhat constrained.

        Now, by allowing a facility to monitor, record, and analyze every phone call placed by a prisoner using AI technology, the government can effectively create a “mosaic” of the inmate’s life. Every call to one’s spouse, children, parents, siblings, or friends can be recorded, analyzed, and stored, allowing the government to piece together intimate details about the lives of prisoners and those with whom they speak on the phone.[54] Although each individual call may provide only a single “mosaic” piece of data, long-term monitoring and recording would allow the government to gather and assemble data about a prisoner over a substantial period of time. Where data collection could “reflect[] a wealth of detail about [one’s] familial, political, professional, religious, and sexual associations,” government abuse is probable.[55] The expansive data collection that is possible with the use of AI-based call monitoring technology, as illustrated by the Mosaic Theory, has implications for prisoners’ First Amendment and substantive due process rights.

        III. Prisoners and the First Amendment

          “[A] prison inmate retains those First Amendment rights that are not inconsistent with his status as a prisoner or with the legitimate penological objectives of the corrections system.”[56] In assessing whether an inmate retains a particular right under the First Amendment, the Court has applied the Turner standard to challenged prison regulations.[57] In effect, this allows prisoners to exercise their First Amendment rights only when doing so does not interfere with the penological interests of prisons. Because the availability of these rights is contingent on the legitimate safety and security needs of prison facilities, courts have not hesitated to uphold prison regulations impinging on traditional First Amendment rights.[58]

          Two notable harms are likely to ensue with the use of this technology. First, the technology’s enhanced ability to perform monitoring and data collection will have a chilling effect on inmates’ desire to communicate over the telephone, greatly restricting one of their only forms of communication with non-prisoner family and friends. Second, because of prison staff’s increased ability to listen in on inmate telephone calls and even record calls based on keyword searches, the potential for retaliatory action against inmates by custody staff is heightened. A discussion on the effects this will have on inmates’ First Amendment rights is important because, other than the limited application of the Eighth and Fourteenth Amendments to protect their rights, the First Amendment is a key constitutional safeguard for incarcerated individuals.[59]

          A. The Chilling Effect on Prisoner Speech

                        As discussed above, advancements in call monitoring technology would allow the government to collect far more data about individuals than was previously possible, raising concerns about individual privacy. In turn, this high level of surveillance is likely to have a chilling effect on speech.[60] “The term ‘chilling effect’ refers to a claim that an otherwise legitimate regulation has the incidental effect of deterring—or chilling—benign activity . . . [like] protected expression.”[61] A chilling effect may also be described as “the spillover effects of laws on benign conduct outside their scope.”[62] Such an effect on speech may give rise to a constitutional claim, as “free speech is an affirmative value, which the government has a particular obligation to promote, or at least not to deter.”[63]

                        “Awareness that the government may be watching chills associational and expressive freedoms.”[64] This “chilling effect” on speech resulting from pervasive surveillance can take effect in two distinct yet related ways.[65] First, people may, on an individual level, choose not to exercise their right to speak based on a fear of privacy harms—for example, potential harm to one’s reputation or the future disclosure of private information collected by the government.[66] These types of harms are described as “affect[ing] the nature of society and imped[ing] individual activities that contribute to the greater social good.”[67] Second, “on a broader societal level,” widespread surveillance chills speech based on a perceived imbalance of power between the government and the people.[68] Where individuals fear reprisal for engaging in certain behaviors or holding certain beliefs, the desire to exercise their rights is inhibited.[69] Thus, government surveillance of speech can affect what information people choose to communicate with each other on an individual level and society’s understanding of the relationship between the government and the people.[70]

            The Supreme Court, in Socialist Workers Party v. Attorney General of the United States, recognized that a plaintiff may have Article III standing to bring a First Amendment challenge based on the possibility of a future “chilling effect” on speech.[71] This marked a shift in the standard required to bring a claim under the First Amendment, as the Supreme Court had previously held that alleging a mere possibility of a general chilling effect was not enough to establish standing.[72] Federal circuits have since recognized that a chilling effect could constitute an abridgment of prisoners’ First Amendment rights.[73] Indeed, “[a] prison inmate retains those First Amendment rights that are not inconsistent with his status as a prisoner or with the legitimate penological objectives of the corrections system.”[74]

            As AI technology advances, though, a person’s status as prisoner has a growing effect on the First Amendment rights prisoners retain. Consider how a facility’s ability to monitor inmate telephone calls would change if it implemented AI-based technology that could monitor and record every call automatically—without requiring staff to sit and listen through hours of telephone calls. Similar issues highlighted by the Mosaic Theory of the Fourth Amendment become apparent in this context. Telephone calls, which serve as an essential form of communication between inmates and people outside the prison walls,[75] can reveal a substantial amount of information about an individual’s life.

            Specifically for prisoners, telephone calls help inmates stay connected with friends and family members, which plays an important role in supporting them during incarceration, and even after, as they are integrated back into the community.[76] Inmates can use the telephone to communicate about what goes on in their daily lives and stay up-to-date on what is happening on the outside, but they may also use it to communicate fears, complaints, or concerns about the conditions of their incarceration. Allowing facilities to monitor and record every inmate telephone call will allow prison administrators to collect more data than was previously possible, effectively enabling them to piece together parts of individual inmates’ lives—both before and during their incarceration—through their telephone communications.

            The U.S. Supreme Court has not explicitly held that inmates have a First Amendment right to use the telephone, and the federal circuits differ on their view of the extent of prisoners’ First Amendment rights.[77] Whether a right to telephone access is recognized or not, that right may not be meaningful where prisons are allowed to conduct mass monitoring and data collection of inmate data. Awareness that the government is listening and recording phone calls constantly will surely have a chilling effect on inmates’ exercise of their First Amendment rights. In this context, it is easy to imagine “the spillover effects of laws on benign conduct outside their scope.”[78] Where they become aware that the government can create a “mosaic” of their life by unnecessarily listening in on harmless conversations, individuals will be deterred from engaging in conversations that reveal personal details about their lives. This, in turn, greatly restricts one of inmates’ most relied-upon forms of communication with non-incarcerated individuals.

            Although incarcerated individuals are subject to “close and continual surveillance,” given the nature of prisons,[79] the Supreme Court has not yet considered the level of surveillance that is possible with newly developed technology; for instance, the ability of prison facilities to monitor and record millions of minutes’ worth of inmate phone calls automatically.[80] Introducing a new method of surveillance that is capable of more extensive and detailed monitoring than previously possible ought to change how courts handle prison surveillance.

            B. A Rise in Staff Retaliation

            A second concern under the First Amendment in this context is an increase in retaliation from staff against inmates based on things they say during telephone calls. Retaliation by prison staff is widespread in facilities across the country.[81] This is due to the imbalance of power between prison inmates and custody staff; as one commentator observed,“[t]he cell door symbolizes surplus power,” and retaliatory actions by prison staff “represent[] an extra-legal use of surplus power.”[82] A claim of retaliation, briefly summed up, “is that a prisoner is penalized for exercising the right of free speech.”[83] Retaliation against an inmate by prison staff can take several forms—moving an inmate to a segregated unit, causing them to lose their job, or spreading rumors about them in the prison population to subject them to harm, assault, or harassment.[84] While the current model of call monitoring does not allow such easy access to listening in on the content of inmates’ phone calls, using AI-based technology would change that. 

            As a result of systemic abuse by prison staff,[85] some inmates use their telephone privileges to share frustrations, complaints, or concerns about their living conditions or custody officers; this is especially true given that inmates often fear retaliation for expressing their dissatisfaction through the prison grievance process.[86] Yet, the AI-based technology’s ability to record conversations based on keyword searches highlights the increased potential for staff abuse or misuse, leading to retaliation. Allowing staff to keyword search inmates’ phone calls enables them to record conversations where inmates discuss frustration with staff members, the conditions of their confinement, or any abuse they may be suffering from.

            In turn, staff can use whatever the inmate said as a basis to retaliate against them, similar to how corrections staff retaliates against inmates for filing grievances against them.[87] The basic idea of retaliation is the same whether the inmate files a grievance or makes a phone call: The inmate expresses dissatisfaction with something related to their incarceration, they make this dissatisfaction known to someone else to seek relief, and ultimately, the fact that the inmate complained is used as a basis for staff to take retaliatory action against them.[88] 

            For a prisoner to bring a successful retaliation claim, he must demonstrate that “(1) he engaged in activity protected by the First Amendment; (2) he suffered a deprivation that would likely deter First Amendment activity in the future; and (3) the First Amendment activity was ‘at least a motivating factor’ in the Defendants’ decision to take the retaliatory action.”[89] However, because there is disagreement between the circuits as to an inmate’s right to use the telephone,[90] an inmate who believes that they are being retaliated against for something they said on the phone may be unable to establish the first prong of this test. At the same time, inmates would have difficulty proving the third prong, as they would have no way of knowing whether the staff was listening in on their phone call. Thus, they could not prove that anything they said on the phone was a motivating factor for adverse action by staff. This is problematic given that abuse and retaliation against inmates is standard behavior in prisons. Among other issues plaguing American prisons, “official misconduct and corruption have left thousands of incarcerated individuals across . . . the nation vulnerable to abuse, assaults, and uncontrolled violence.”[91]

            In light of the chilling effects on speech and the increased likelihood of staff retaliation and abuse, automatic call monitoring presents severe risks to prisoners’ First Amendment rights. As outlined in Part V, the security benefits of AI technology do not outweigh inmates’ constitutional rights under the Turner test.

            IV. Prisoners and Substantive Due Process

                          Aside from the possible First Amendment harms to prisoners when their phone calls are monitored using AI technology, pervasive monitoring and data collection would do away with any liberty interests retained by prisoners in keeping certain details of their lives private.

              Substantive due process is best understood as a concept within constitutional law “ensuring that essential personal freedoms are protected from unwarranted government intrusion.”[92] The basis of substantive due process is derived from several constitutional amendments that “refer to or assume the existence of such rights,” although they are not explicitly included in the Bill of Rights.[93] The rationale for inferring such rights is based on the notion that the government should not be allowed to infringe on certain rights, even if they are not enumerated in the Constitution itself, that are “essential to the orderly pursuit of happiness by free men.”[94] This “orderly pursuit of happiness” stems from the text of the Fifth and Fourteenth Amendments themselves, proscribing that “[n]o person shall . . . be deprived of life, liberty, or property, without due process of law.”[95] Because these rights are not enshrined in the Constitution itself, courts must rely on decisions of the Supreme Court recognizing that certain rights are “‘implicit in the concept of ordered liberty,’ such that ‘neither liberty nor justice would exist if they were sacrificed.’”[96]

              Over the past century, the Court has recognized various distinct rights that fall within this category—one of the most essential being privacy.[97] As described by Justice Douglas in Griswold v. Connecticut, a famous substantive due process rights decision, “specific guarantees in the Bill of Rights have penumbras, formed by emanations from those guarantees that help give them life and substance.”[98] Several enumerated rights in the Constitution imply that there is a right to privacy within the concept of due process.[99] The First Amendment protects the right of association; the Third Amendment prohibits the quartering of soldiers in one’s home; the Fourth Amendment protects individuals from unreasonable government interference with their persons, houses, papers, or effects; the Fifth Amendment “enables the citizen to create a zone of privacy which government may not force him to surrender to his detriment;” and the Ninth Amendment prohibits a construction of the constitution that would deny or disparage the rights of the people.[100] Together, these amendments suggest that one’s right to privacy is so “implicit in the concept of ordered liberty” that an individual may not be deprived of their privacy without due process of law.[101]

              For incarcerated individuals, the Supreme Court has ruled that “[p]risoners may also claim the protections of the Due Process Clause. They may not be deprived of life, liberty, or property without due process of law.”[102] The Court in Wolff v. McDonnell stressed that although imprisonment “necessarily makes unavailable many rights and privileges of the ordinary citizen . . . a prisoner is not wholly stripped of constitutional protections when he is imprisoned for crime.”[103] Prisoners have protected liberty interests in the revocation of probation[104] or parole,[105] as well as the revocation of good-time credits granted by a state.[106] The Supreme Court has held that prisoners also retain other narrowly defined liberty interests during their incarceration.[107]

              Although prisoners have no expectation of privacy in their cells,[108] some federal circuits have held that an inmate ‘“maintains some reasonable expectations of privacy while in prison . . . even though those privacy rights may be less than those enjoyed by non-prisoners.”’[109] However, several other circuits, including the Seventh Circuit, hold that prisoners generally do not have constitutionally protected expectations of privacy.[110] Despite these differences in the federal circuits’ recognition of inmate privacy interests, “a prison inmate retains those constitutional rights that are not inconsistent with his status as a prisoner or with the legitimate penological objectives of the corrections system.”[111]

              Through a series of decisions, the Court has made clear that the due process clause applies to prisoners, regardless of their incarceration status, where they have retained liberty interests.[112] Thus, inmates arguably retain a liberty interest in keeping private from the government the intimate details of their lives that can be protected without interfering with any penological objectives of the corrections system, even though no court has explicitly held this.[113] Although it is undisputed that prisoner telephone calls can be monitored for facility security purposes,[114] courts have never conferred on prison administrators a right to collect and store unlimited inmate data.

              Compared to the ability of prisons to monitor inmate phone calls currently, the automatic monitoring and recording of every single phone call would mark a dramatic shift in prisons’ collection of inmate data. A facility may have an interest in monitoring inmate phone calls to protect facility security by intercepting contraband trafficking and preventing future incidents of violence, but it is not clear what penological interest, if any, would be served by maintaining a comprehensive profile of an inmate’s life. This technology has the potential to infringe on inmates’ liberty interest in maintaining some privacy in the details of their lives, separate from their privacy rights forfeited incident to their incarceration.

              V. The Case Against Artificial Intelligence in Prisons

              In applying the Turner test to prisons’ use of this technology, AI-based surveillance technologies cannot be justified in light of the rights at stake. As the Court laid out in Turner, four factors guide the analysis of whether a prison regulation is reasonable: (1) whether there is a valid, rational connection between the prison regulation and the governmental interest justifying it; (2) whether there are alternative means available for prisoners to exercise the right; (3) whether the accommodation of the asserted right will have a “ripple effect” on inmates and staff; and (4) whether there is a “ready alternative” that fully accommodates the prisoners’ rights without interfering with valid penological interests.[115] The first prong is harder to challenge, as prison administrators undisputedly have an interest in monitoring the calls of inmates for facility security reasons. The other three factors, however, weigh in favor of preserving inmates’ liberty interests.

              As for the second factor, prisoners’ ability to communicate with non-incarcerated individuals is generally limited to in-person visitations, telephone calls, and communication through letters. However, because of the chilling effect that such a use of technology would likely have on prisoners, inmates may be more likely to rely on visitations and letters for communicating with friends and family. In some states, like Indiana, visitations are a privilege, not a right, and visitation privileges are subject to restrictions based on disciplinary conduct or housing status.[116] This leaves prisoners communicating via letters, which is also subject to pervasive monitoring by prison staff and is far less efficient than communicating by phone call.[117]

              Bearing in mind the importance of prisoners maintaining their ability to communicate with loved ones regularly,[118] arguing that communicating by letter is an adequate alternative to telephones is unconvincing in the twenty-first century. As the Bureau of Prisons has acknowledged, “telephone privileges are a necessary benefit to keep inmates in a cooperative mode.”[119] If the telephone privileges are to be meaningful, inmates deserve assurance that their private calls will only be monitored and recorded where there are legitimate security concerns—not at the will of prison staff who may have improper motives. 

              Regarding the third factor, simply prohibiting the use of AI to monitor calls would be the “accommodation” of the right sought in this scenario. As such, accommodating the right would not have a significant ripple effect on the facility because nothing would change from the current practice of call monitoring. In addition, this note highlighted the significant chilling effect that is likely if the right is not accommodated. As Justice Sotomayor stated in her concurrence in United States v. Jones, “[a]wareness that the government may be watching chills associational and expressive freedoms.”[120]

              Given the nature of their incarceration, prisoners already retain almost no privacy in their daily lives. To preserve even the slightest bit of privacy, inmates’ desire to use their telephone privileges for communication with friends and family would surely be deterred. Otherwise, the state has the authority to know everything about the inmate’s life while in the state’s custody, plus everything that he communicates to a spouse, sibling, parent, or friend during any phone conversation, whether threatening or merely innocuous.

              The fourth factor weighs in favor of preserving inmates’ rights, too, as there is a ready alternative to monitoring inmate calls without interfering with penological interests that facilities already employ: relying on corrections staff to monitor phone calls. Courts give prisons great deference in maintaining facility security.[121] As such, they are free to devote as many staff resources as desired to monitor inmate phone calls. The fact that many prisons in the United States are understaffed cannot entitle the government to further erode inmate rights;[122] if this were the result, prisoners’ rights would continue to be vulnerable to prison staffing levels. Prison facilities have relied on corrections staff to monitor prisoner phone calls since the 1970s, when the Bureau of Prisons first established telephone access programs for inmates.[123] Until prisoners’ privacy is better protected in light of new surveillance technology, prison administrators should not be able to further erode prisoners’ substantive due process rights by monitoring, recording, and storing every minute of their telephone calls.

              Thus, although there is a legitimate penological interest in monitoring inmates’ telephone calls for facility security, the possible intrusions on inmates’ First Amendment and substantive due process rights resulting from the use of AI-driven call monitoring technology cannot be justified under the Turner standard. As a practical matter, telephone calls are an extremely important form of communication both for inmates to stay in touch with family and friends and for facilities to gain compliance from inmates.[124]  Employing technology that discourages inmates’ use of the telephone harms both inmates and prison administrators. And as a constitutional matter, Justice White clearly enunciated in Wolff v. McDonnell that “[t]here is no iron curtain drawn between the Constitution and the prisons of this country.”[125] Although prisons should be afforded some latitude to ensure facility security, “a prisoner is not wholly stripped of his constitutional protections when he is imprisoned,” including First Amendment and substantive due process protections.[126]

              Conclusion

              Given that several states have begun implementing AI-based technology for improving facility security, others may follow the movement of implementing AI in corrections. However, despite the technology’s utility in detecting speech during telephone conversations that may pose safety and security threats to prison facilities, there are severe drawbacks to its use. By analogizing to the Mosaic Theory of the Fourth Amendment, this note highlighted the possible First Amendment and substantive due process concerns associated with advanced data collection technology.

              Ultimately, considering that inmates’ rights will be further eroded where AI is used for call monitoring, this Note argues that using AI for this function is not justified under the Supreme Court’s test in Turner.[127] Undoubtedly, prisons should seek to improve safety and security, but prisoners retain certain rights such that “[w]hen a prison regulation or practice offends a fundamental constitutional guarantee, federal courts [must] discharge their duty to protect constitutional rights.”[128] This is especially true because, as stressed throughout this note, “a prisoner is not wholly stripped of constitutional protections when he is imprisoned for a crime.”[129] Allowing prisons to collect such a vast amount of prisoner data causes a further erosion of prisoners’ retained First Amendment and substantive due process rights, which is an unacceptable result for an already vulnerable class of citizens.[130]  


              [1] Kentrell Owens, Camille Cobb & Lorrie Faith Cranor,‘You Gotta Watch What You Say’: Surveillance of Communication with Incarcerated People, Ass’n for Computing Mach. 2–3 (May 7, 2021), https://dl.acm.org/doi/10.1145/3411764.3445055 [perma.cc/L988-2GCD].

              [2] Criminal Calls: A Review of the Bureau of Prisons’ Management of Inmate Telephone Privileges, U.S. Dep’t of Just. & Off. of the Inspector Gen., Executive Summary (Aug. 1999), https://oig.justice.gov/sites/default/files/archive/special/9908/index.htm [perma.cc/L2EC-VR6R].

              [3] See, e.g., Telephone Privileges, Ind. Dep’t of Corr. (Nov. 1, 2023), https://www.in.gov/idoc/files/policy-and-procedure/policies/02-01-105-Telephones-11-1-2023.pdf [perma.cc/BKF8-66PQ].

              [4] Joe Russo, Dulani Woods, John S. Shaffer & Brian A. Jackson, Countering Threats to Correctional Institution Security, Rand Corp. (2019), https://www.rand.org/content/dam/rand/pubs/research_reports/RR2900/RR2933

              /RAND_RR2933.pdf [perma.cc/6WEP-3249].

              [5] Chris Francescani, US prisons and jails using AI to mass-monitor millions of inmate calls, Abc News (Oct. 24, 2019), https://abcnews.go.com/Technology/us-prisons-jails-ai-mass-monitor-millions-inmate/story?id=66370244 [perma.cc/QFY5-ZUVB].

              [6] Here’s What You Need to Know About Leo Technologies and Verus, Leo Tech. 1, https://media.businesshumanrights.org/media/documents/files/documents/060920_LeoTech_FactSheet.pdf [perma.cc/76FV-SK5K] (last visited Feb. 7, 2026).

              [7] Grayce Burns, The use and future of artificial intelligence monitoring in prisons, Reason Found. (Feb. 1, 2022), https://reason.org/commentary/the-use-and-future-of-artificial-intelligence-monitoring-in-prisons/ [perma.cc/89VY-GKMA].

              [8] Niclas Rautenberg & Daragh Murray, Making Tangible the Long‑Term Harm Linked to the Chilling Effects of AI‑enabled Surveillance: Can Human Flourishing Inform Human Rights?, 25 Hum. Rts. Rev. 293, 296 (2024).

              [9] Thornburgh v. Abbott, 490 U.S. 401, 407 (1989) (quoting Turner v. Safley, 482 U.S. 78, 84 (1987)).

              [10] Turner, 482 U.S. at 89.

              [11] Id. at 81.

              [12] Id. at 89.

              [13] Id.

              [14] Id. at 90.

              [15] Id.

              [16] Id.

              [17] Id.

              [18] U.S. Dep’t of Just. & Off. of the Inspector Gen., supra note 2, at Ch. 7.

              [19] Id. 

              [20] Arsberry v. Illinois, 244 F.3d 558, 564–65 (7th Cir. 2001).

              [21] Boriboune v. Litscher, 91 F. App’x 498, 499 (7th Cir. 2003).

              [22] U.S. Dep’t of Just. & Off. of the Inspector Gen., supra note 2, at Ch. 7; Turner, 482 U.S. at 81.

              [23] Rashi Maheshwari, Advantages Of Artificial Intelligence (AI) in 2025, Forbes (Aug. 24, 2023, at 13:12 ET), https://www.forbes.com/advisor/in/business/software/advantages-of-ai/ [perma.cc/GGC3-SFUH].

              [24] See All the Benefits of Artificial Intelligence, W. Governors Univ. (Apr. 25, 2022), https://www.wgu.edu/blog/benefits-artificial-intelligence2204.html#:~:text=Medical

              %20Progression,AI%20does%20the%20busy%20work. [perma.cc/78VV-N52U].

              [25] Steven Schuetz et al., Artificial Intelligence Applications in Corrections, U.S. Dep’t of Just., Nat’l Inst. of Just. 7 (Aug. 2020), https://cjtec.org/files/5f5f9458ebc72 [perma.cc/HC8N-WU7N].

              [26] Id. 

              [27] Francescani, supra note 5.

              [28] U.S. Dep’t of Just. & Off. of the Inspector Gen., supra note 2, at Ch. 1.

              [29] Shannon Heffernan & Weihua Li, As prison populations rise, states face a stubborn staffing crisis, Usa Today (Jan. 10, 2024, at 6:09 ET), https://www.usatoday.com/story/news/nation

              /2024/01/10/new-data-dire-state-prison-staffing-shortage/72166165007/ [perma.cc/9PZ7-3GQ5].

              [30] Francescani, supra note 5.

              [31] Id. 

              [32] Debra Cassens Weiss, Prisons and jails use artificial intelligence to monitor inmate phone calls, ABA J. (Oct. 25, 2019, at 10:22 CDT), https://www.abajournal.com/news/article/prisons-and-jails-use-artificial-intelligence-to-monitor-inmate-phone-calls [perma.cc/YA3L-9UDS].

              [33] Stephanie Kanowitz, AI on the Line: Monitoring prisoners’ phone calls for criminal intent, Route Fifty (Aug. 20, 2021), https://www.route-fifty.com/infrastructure/2021/08/ai-on-the-line-monitoring-prisoners-phone-calls-for-criminal-intent/316135/ [perma.cc/5V86-X7RE].

              [34] Justin Garcia, Florida prisons use artificial intelligence to surveil calls, Tampa Bay Times (Sep. 19, 2023), https://www.tampabay.com/news/florida/2023/09/19/florida-artificial-intelligence-prison-surveillance-leo-technologies-verus-calls-amazon/ [perma.cc/Z4SL-BK5R].

              [35] Leo Tech., supra note 6, at 1.

              [36] Id. 

              [37] Id. 

              [38] Id. at 2.

              [39] See Orin S. Kerr, The Mosaic Theory of the Fourth Amendment, 111 Mich. L. Rev. 311, 320 (2012).

              [40] United States v. Jones, 565 U.S. 400, 427 (2012) (Alito, J., concurring).

              [41] Robert Fairbanks, Masterpiece or Mess: The Mosaic Theory of the Fourth Amendment Post-Carpenter, 26 Berkeley J. Crim. L. 71, 74 (2021).

              [42] Jones, 565 U.S. at 402.

              [43] See id. at 429 (Alito, J., concurring); id. at 416 (Sotomayor, J., concurring).

              [44] Id. at 429 (Alito, J., concurring).

              [45] See id. at 403 (noting that law enforcement obtained over 2,000 pages of suspect’s location data using a GPS tracker).

              [46] Id. at 429 (Alito, J., concurring).

              [47] Id. at 416 (Sotomayor, J., concurring) (citation omitted).

              [48] Id.

              [49] Fairbanks, supra note 41, at 74.

              [50] Id.

              [51] See Hudson v. Palmer, 468 U.S. 517, 525–26 (1984).

              [52] Id.

              [53] See United States v. Sababu, 891 F.2d 1308, 1329 (7th Cir. 1989).

              [54] The State of Prison & Jail Communication Systems, Nat’l Ass’n of Crim. Def. Laws. (Mar. 21, 2022), https://www.nacdl.org/Map/State-of-Prison-Jail-Call-Communication-Systems#:~:text=The%20three%20most%20commonly%20available,%2C%20and%2024%25%20 [https://perma.cc/Q5V9-KWLB%5D.

              [55] United States v. Jones, 565 U.S. 400, 415–16 (2012) (Sotomayor, J., concurring).

              [56] Pell v. Procunier, 417 U.S. 817, 822 (1974).

              [57] Prison Free Speech and Government as Prison Administrator, Const. Ann. Amend. 1.7.8.5, https://constitution.congress.gov/browse/essay/amdt1-7-8-5/ALDE_00000758/ [https://perma.cc/5UKW-H8HT] (last visited Feb. 12, 2026).

              [58] See, e.g., Jones v. N. C. Prisoners’ Lab. Union, Inc., 433 U.S. 119, 136 (1977) (upholding restriction on the operation and solicitation of unions within prison); Beard v. Banks, 548 U.S. 521, 525 (2006) (upholding restriction on prisoners’ reading materials including newspapers and magazines); Thornburgh v. Abbott, 490 U.S. 401, 403–04 (1989) (upholding restriction that excluded certain incoming publications and correspondence).

              [59] Kevin Francis O’Neill, Rights of Prisoners, Free Speech Ctr. (July 9, 2024) https://firstamendment.mtsu.edu/article/rights-of-prisoners/ [https://perma.cc/PYQ5-RGJP%5D.

              [60] See Paulina Perlin, ACLU v. NSA: How Greater Transparency Can Reduce the Chilling Effects of Mass Surveillance, Media Freedom & Info. Access Clinic (Dec. 6, 2017), https://law.yale.edu/mfia/case-disclosed/aclu-v-nsa-how-greater-transparency-can-reduce-chilling-effects-mass-surveillance#No11 [perma.cc/G94D-QCAA] (“The chilling effects of mass government surveillance programs are well demonstrated—and even quantified—in scholarly literature and case law alike.”) (citations omitted).

              [61] Leslie Kendrick, Speech, Intent, and the Chilling Effect, 54 Wm. & Mary L. Rev. 1633, 1649 (2013).

              [62] Id.  

              [63] Id. at 1655.

              [64] United States v. Jones, 565 U.S. 400, 416 (2012) (Sotomayor, J., concurring).

              [65] Jonathon W. Penney, Understanding Chilling Effects, 106 Minn. L. Rev. 1451, 1480–81 (2022).

              [66] Id.

              [67] Daniel J. Solove, A Taxonomy of Privacy, 154 U. Pa. L. Rev. 477, 488 (2006).

              [68] Penney, supra note 65, at 1480.

              [69] Id. 

              [70] Id. 

              [71] 419 U.S. 1314, 1319 (1974).

              [72] Laird v. Tatum, 408 U.S. 1, 13–14 (1972) (quoting United Pub. Workers of Am. (C.I.O.) v. Mitchell, 330 U.S. 75, 89 (1947)) (“Allegations of a subjective ‘chill’ are not an adequate substitute for a claim of specific present objective harm or a threat of specific future harm; ‘the federal courts established pursuant to Article III of the Constitution do not render advisory opinions.’”).

              [73] See, e.g., Gassler v. Wood, 14 F.3d 406, 408–09 (8th Cir. 1994) (“If [plaintiffs’] First Amendment communicative rights were abridged at all, it would be because of the deterrent or ‘chilling’ effect upon free expression.”); O’Keefe v. Van Boening, 82 F.3d 322, 325 (9th Cir. 1996) (“Unconstitutional restrictions may arise from the deterrent, or ‘chilling,’ effect of governmental action.”); Haze v. Harrison, 961 F.3d 654, 658 (4th Cir. 2020) (“Opening an incarcerated person’s legal mail outside of his presence can chill protected speech.”); Jones v. Brown, 461 F.3d 353, 359 (3d Cir. 2006) (holding that a facility’s practice of opening legal mail “deprives the expression of confidentiality and chills the inmates’ protected expression”).

              [74] Pell v. Procunier, 417 U.S. 817, 822 (1974).

              [75] Nat’l Ass’n of Crim. Def. Laws., supra note 54.

              [76] Ryan Shanahan & Sandra Villalobos Agudelo, The Family and Recidivism, Vera Inst. of Just. (Sep.–Oct. 2012), https://www.prisonpolicy.org/scans/vera/the-family-and-recidivism.pdf [https://perma.cc/3P4H-KT4N%5D.

              [77] Peter R. Shults, Calling the Supreme Court: Prisoners’ Constitutional Right to Telephone Use, 92 B.U. L. Rev. 369, 380 (2012).

              [78] Kendrick, supra note 61, at 1649.

              [79] Hudson v. Palmer, 468 U.S. 517, 527–28 (1984).

              [80] See Avi Asher-Schapiro & David Sherfinski, Crack down on U.S. prison surveillance tech, rights groups urge, Thomson Reuters Found. (Feb. 10, 2022, at 15:30 GMT), https://news.trust.org/item/20220210152812-a16ki/ [https://perma.cc/QSE9-9S8D%5D.

              [81] See James E. Robertson, “One of the Dirty Secrets of American Corrections”: Retaliation, Surplus Power, and Whistleblowing Inmates, 42 U. Mich. J.L. Reform 611, 613 (2009).

              [82] Id. at 615.

              [83] Farrow v. West, 320 F.3d 1235, 1248 (11th Cir. 2003) (quoting Thomas v. Evans, 880 F.2d 1235, 1242 (11th Cir. 1989)).

              [84] Kristofer Seneca, ‘No One’s Going to Believe Them’, Prison Journalism Project (Jan. 30, 2024) https://prisonjournalismproject.org/2024/01/30/corrections-officers-abuse-inmates-common/ [https://perma.cc/9RCM-V5D4%5D; Robertson, supra note 81, at 613–14.

              [85] Seneca,supra note 84.

              [86] Melissa Benerofe, Collaterally Attacking the Prison Litigation Reform Act’s Application to Meritorious Prisoner Civil Litigation, 90 Fordham L. Rev. 141, 160 (2021).

              [87] See Roberston, supra note 81, at 613–14.

              [88] See Retaliation Against Prisoners for Protected First Amendment Expression, Ams. for Effective L. Enf’t 2–4 (Mar. 2010), https://www.aele.org/law/2010all03/2010-03MLJ301.pdf [https://perma.cc/Y6ZZ-5T9H]. 

              [89] Bridges v. Gilbert, 557 F.3d 541, 546 (7th Cir. 2009).

              [90] See supra notes 18–21 and accompanying text.  

              [91] Prison Conditions, Equal Just. Initiative, https://eji.org/issues/prison-conditions/ [https://perma.cc/4KNH-3HXS%5D (last visited Dec. 18, 2025).

              [92] Substantive Due Process, Legal Info. Inst. (June 2024), https://www.law.cornell.edu/wex/substantive_due_process [https://perma.cc/F88Y-7643%5D.

              [93] Id. 

              [94] Meyer v. Nebraska, 262 U.S. 390, 399 (1923).

              [95] U.S. Const. amend. V; U.S. Const. amend. XIV, §1. 

              [96] Washington v. Glucksberg, 521 U.S. 702, 721 (1997) (citations omitted).

              [97] Griswold v. Connecticut, 381 U.S. 479, 484 (1965).

              [98] Id. 

              [99] Id. 

              [100] Id.

              [101] Id. at 500 (Harlan, J., concurring) (quoting Palko v. Connecticut, 302 U.S. 319, 325 (1937)); see also Meyer v. Nebraska, 262 U.S. 390, 399 (1923).

              [102] Wolff v. McDonnell, 418 U.S. 539, 556 (1974).

              [103] Id. at 555. 

              [104] Morrissey v. Brewer, 408 U.S. 471, 483 (1972).

              [105] Gagnon v. Scarpelli, 411 U.S. 778, 782 (1973).

              [106] Wolff, 418 U.S. at 557.

              [107] See, e.g., Vitek v. Jones, 445 U.S. 480, 493 (1980) (“We conclude that a convicted felon also is entitled to the benefit of procedures appropriate in the circumstances before he is found to have a mental disease and transferred to a mental hospital.”); Washington v. Harper, 494 U.S. 210, 221–22 (1990) (“We have no doubt that . . . respondent possesses a significant liberty interest in avoiding the unwanted administration of antipsychotic drugs under the Due Process Clause of the Fourteenth Amendment.”).

              [108] Hudson v. Palmer, 468 U.S. 517, 525–26 (1984).

              [109] Stoudemire v. Mich. Dep’t of Corr., 705 F.3d 560, 572 (6th Cir. 2013) (quoting Cornwell v. Dahlberg, 963 F.2d 912, 916 (6th Cir. 1992)); see also Hutchins v. McDaniels, 512 F.3d 193, 196 (5th Cir. 2007) (“[T]he Fourth Amendment protects prisoners from searches and seizures that go beyond legitimate penological interests. Searches of prisoners must be conducted in a manner that is reasonable under the facts and circumstances in which they are performed.”) (citations omitted). 

              [110]  Mathews v. Raemisch, 513 F. App’x 605, 608 (7th Cir. 2013); United States v. Balon, 384 F.3d 38, 44 (2d Cir. 2004); Washington v. Grace, 445 F. App’x 611, 615–16 (3d Cir. 2011).

              [111] Turner v. Safley, 482 U.S. 78, 95 (1987) (quoting Pell v. Procunier, 417 U.S. 817, 822 (1974)).

              [112] See supra notes 102–07 and accompanying text.

              [113] See Jennifer A. Brobst, The Metal Eye: Ethical Regulation of the State’s Use of Surveillance Technology and Artificial Intelligence to Observe Humans in Confinement, 55 Cal. W. L. Rev. 1, 33 (2018) (“The protection of autonomy, even for those subjected to surveillance in state institutions, finds support in common law, the federal penumbra of privacy rights, and a growing number of express privacy provisions in state constitutions across the United States.”).

              [114] See Washington v. Meachum, 630 A.2d 262, 275–76 (Conn. 1996) (collecting cases).

              [115] Turner v. Safley, 482 U.S. 78, 89–91 (1987).

              [116] Visitation Policy Summary, Ind. Dep’t of Corr. Pol’y Statement, https://www.in.gov/idoc/facilities/supporthub/visitation/#:~:text=An%20offender%20found%20guilty%20in,unable%20to%20travel%20to%20facilities [https://perma.cc/P27V-RN9U] (last visited Dec. 18, 2025).

              [117] See Correspondence, Policy and Administrative Procedure, Ind. Dep’t of Corr. 12–13 (Aug. 2023), https://www.in.gov/idoc/files/policy-and-procedure/policies/02-01-103-Correspondence-8-22-2023.pdf [https://perma.cc/BQ9S-E5W5%5D.

              [118] Shanahan & Villalobos Agudelo, supra note 76.

              [119] U.S. Dep’t of Just. & Off. of the Inspector Gen., supra note 2, at Ch. 7.

              [120] United States v. Jones, 565 U.S. 400, 416 (2012) (Sotomayor, J., concurring).

              [121] Turner v. Safley, 482 U.S. 78, 84–85 (1987).

              [122] Erica Bryant, Corrections Staffing Shortages Offer Chance to Rethink Prison, Vera Inst. of Just. (Nov. 1, 2024) https://www.vera.org/news/corrections-staffing-shortages-offer-chance-to-rethink prison#:~:text=Prisons%20across%20the%20country%20are,lockdowns%20are%20becoming%20the%20norm. [https://perma.cc/VZ2S-CYNC%5D

              [123] U.S. Dep’t of Just. & Off. of the Inspector Gen., supra note 2, at Ch. 2.

              [124] Id. at Ch. 7.

              [125] Wolff v. McDonnell, 418 U.S. 539, 555–56 (1974).

              [126] Id. at 555. 

              [127] Turner v. Safley, 482 U.S. 78, 89 (1987).

              [128] Id. at 84 (quoting Procunier v. Martinez, 416 U.S 396, 405–06 (1974)).  

              [129] Wolff, 418 U.S. at 555.

              [130] Justin Iverson, Surveilling Potential Uses and Abuses of Artificial Intelligence in Correctional Spaces, 9 Lincoln Mem’l U. L. Rev. (Summer 2022), at 4.

              Leave a comment