Digital Justice or Digital Divide? The Role of Technology in Bail Reform and Access to Justice for Undertrial Prisoners in India

Paper Code: AIJACLAV12RP2026

Category: Research Paper

Date of Publication: Sep 10, 2026

Citation: Ms. Disha S Patil, & Mr. Prashik Rangari, “Digital Justice or Digital Divide? The Role of Technology in Bail Reform and Access to Justice for Undertrial Prisoners in India", 6, AIJACLA, 94-110 (2026), <https://www.aequivic.in/post/digital-justice-or-digital-divide-the-role-of-technology-in-bail-reform-and-access-to-justice-for-u>

Author Details: Ms. Disha S Patil, & Mr. Prashik Rangari,

Students of 6th Sem BALLB, Christ Academy Institute of Law, Bengaluru

Download PDF

Abstract

The criminal justice system in India currently experiences an extreme undertrial crisis, resulting in extended pretrial detention that violates the constitutional presumption of innocence and the right to personal freedom established under Article 21 of the Constitution of India. A substantial proportion of accused persons remain incarcerated while awaiting trial due to judicial backlogs, procedural delays, and socio-economic barriers that prevent them from fulfilling bail conditions. The situation creates major problems that affect people's ability to obtain justice and the fairness of criminal processes and their essential rights.

The Indian judiciary has started using digital technologies to enhance efficiency, transparency, and accessibility. This study examines how technological solutions improve bail procedures while enabling remote court proceedings to become more accessible and providing essential legal knowledge to undertrial inmates throughout the criminal justice system.

The development of digital justice programs creates two essential problems that need to be solved. The combination of inadequate digital skills, lack of essential infrastructure, and limited access to digital resources creates a digital barrier that prevents people from reaching the legal system, especially those who belong to marginalized communities and who experience economic difficulties. The paper examines how technology serves two functions in bail reform while it investiates how new problems arise from using technology, which includes problems with algorithmic discrimination and the need for transparent procedures and accountable systems when artificial intelligence is used in judicial settings. The study determines whether the digital changes occurring in India’s criminal justice system will deliver actual digital justice to undertrial prisoners or whether they will further strengthen exigsting social power imbalances.

Keywords: Undertrial Prisoners, Bail Reform, Access to Justice,  Digital Justice, Digital Divide, Artificial Intelligence in Judiciary.

Paper Code: AIJACLAV12RP2026

1. Introduction

Undertrial prisoners are those persons against whom an offense has been registered, but they have not yet been convicted and found guilty by the court of law. They are kept in judicial custody, and they are presumed innocent. The undertrial prisoners are not given the same kind of sentence as is given to the convicts. However, they are kept in custody for a very long time because of the lack of proper legal representation and because of other socioeconomic factors.

The undertrial crisis in India’s criminal justice system is acute and questions basic human liberty and access to justice. According to the statistics released by the National Crime Records Bureau (NCRB) for the year 2023, the total percentage of undertrial prisoners is 74%, which is approximately [1]3.9 lakh out of the total 5.3 lakh inmates in the country. Although this is an improvement over the previous years, as in 2021 the total percentage was 77%, and in 2022, the total percentage was 76%, the percentage is still very high. Moreover, the problem is complicated by the fact that the backlog is very high, as there are a number of cases pending in the courts, and the ratio of judges to the total population is very low, especially for the economically weaker section of society. Therefore, the undertrial prisoners are kept in custody for months and years before they are actually convicted, and this is actually considered to be punishment in itself.

This scenario is related to [2]Article 21 of the Constitution of India, as the Constitution guarantees the fundamental right to life and liberty. The judiciary has also given the right to a speedy trial, and this is done to ensure that the person is not kept in custody for an unnecessarily long time.[3]

 The constitutional requirement becomes more important because the Bharatiya Nagarik Suraksha Sanhita (BNSS) 2023[4], which replaces the CrPC, establishes new rules for using technology in court trials and for handling first-time offenders. The constitutional guarantee faces serious threats because of ongoing delays throughout the system and the difficulties associated with putting the new digital requirements into practice.

As a consequence of the aforementioned challenges, there has been an increasing adoption of digital technologies in the Indian judiciary, including virtual hearings, e-courts, and integrated case management, which are geared towards the efficient and accessible nature of the judicial process. The adoption of digital technologies in the judiciary has led to the concept of ‘digital justice,’ which has been explained as the pursuit of an efficient, transparent, and accessible legal process. At the same time, the adoption of digital technologies has created an inherent tension in the sense that, while there is the potential to enhance access to the judicial process, the concept of the digital divide has been created.

The digital reforms fail to provide advantages to undertrial prisoners who come from socio-economically disadvantaged communities because they cannot access the necessary resources. The system creates a paradox because the technology developed to improve justice access for all people actually prevents access for those who would benefit from it the most. The primary issue that needs to be solved is whether digital transformation in the judiciary promotes inclusive practices or establishes new methods of exclusion.

1.1 Objectives of the Study:

1.       To examine and investigate how digital technology facilitates bail processes, which enable people to access justice.

2.       To evaluate how digital technology disparities affect the rights of undertrial prisoners.

3.       To analyze whether digitalization leads to fair or unfair distribution of justice services.

1.2 Research Questions :

1. To what extent does digitalization affect indigent undertrial prisoners' access to bail?

2 How does the digital divide create new barriers to justice for marginalized sections of the prison population?

3. The research aims to discover methods for constructing digital systems that will deliver justice services in an equitable manner to all individuals.

2. Conceptual Framework & Methodology

2.1 Digital Justice

Digital justice refers to the use of digital technologies to improve the operations of courts in their pursuit of justice. The system requires electronic case management systems, e-filing platforms, virtual court hearings, and online legal information resources to operate efficiently. The technologies create a new way for courts to function by enabling better public access and judicial work to progress at a greater speed and with greater visibility.

In particular, the use of digital technology allows for the expeditious disposal of cases through the simplification of documentation, real-time tracking of cases, and improved communication between judges, court personnel, and parties. Digital case management also allows for data collection, which in turn aids evidence-based decision-making, the judicious allocation of judicial time, and the identification of procedural bottlenecks[5]. At the same time, the digital recording and dissemination of judicial proceedings enhances the transparency and accountability of the judicial process.

About the Indian experience, the concept of digital justice has assumed prominence through the use of e-courts, virtual courts, and the digitalization of judicial records. . This is now supported by the Bharatiya Nagarik Suraksha Sanhita (BNSS), 2023, which has established the statutory framework for the conduct of trials electronically and the transmission of summons electronically, thus theoretically bridging the distance between physical distance and judicial presence. However, the success of such an integration of technology will depend on Human-Centric Design to ensure the system remains accessible to the layperson.

2.2 DIGITAL DIVIDE:

The term 'digital divide' refers to the gap between individuals with access to digital technologies and those without, as well as the gap in the proficiency of utilizing these technologies[6]. This is not limited to the level of access to technology and the internet; rather, it also extends to the level of proficiency and trust in these technologies.

The digital divide exists on various levels, including limited infrastructure, inadequate skills, and insufficient awareness of how to effectively use these technologies. In the context of the justice sector, this might hinder the ability of certain individuals to benefit from the digital justice movement. Marginalized communities, the less affluent, and rural communities are the ones that are likely to be affected due to limited access to technology and lower levels of digital proficiency.

The phenomenon of the digital divide assumes particular significance in the context of undertrial prisoners, as these individuals are likely to be socio-economically disadvantaged classes of individuals and are likely to have limited knowledge of the law and limited access to technology.

2.3 Digital Justice and the Digital Divide: The Emerging Paradox

Digital justice aims to make the legal system more efficient, transparent, and accessible. However, the issue of the digital divide acts as a significant barrier to the actualization of the concept of digital justice. This is because the same technologies that can bring democracy to the legal system can also widen the existing inequalities if not carefully implemented.

The adoption of digital justice can make the legal system more efficient for those who have access to and the ability to utilize the relevant technologies. This means that those who do not have access to and the ability to utilize the relevant technologies can become more excluded in the legal system than before. This implies that while the legal system can become more efficient for some due to the adoption of digital justice, it can also become less efficient for others. This emerging paradox is the main focus of this study. It reveals that while assessing the impact of the adoption of digital justice in the legal system, one should not only consider its ability to make the system more efficient but also more inclusive.

This study aims to examine whether the adoption of digital technologies in the bail process can make the legal system more accessible to undertrial prisoners and more efficient in providing justice to them, or if it can widen the existing inequalities.

2.4 Methodology

A mainly doctrinal research methodology is used for this study. This means that there is an emphasis on analysis and interpretation of relevant statutes and judicial pronouncements to assess the role of digital technologies in bail and access to justice for undertrial prisoners in India. For instance, relevant legislative frameworks such as the Bharatiya Nyaya Suraksha Sanhita, 2023, and judicial pronouncements and reports published by relevant government departments and institutions such as the National Crime Records Bureau (NCRB) are used for this purpose.

An analytical research perspective is used to assess the practical implications of digitalization in the justice system and its role in addressing socio-economic disparities and access to justice for undertrial prisoners in India.

3. Legal and Constitutional Framework

The legal and constitutional framework on bail and undertrial prisoners’ rights has developed significantly in India, and while the right to liberty under Article 21 of the Constitution of India serves as a foundation, its implementation has largely depended on how the Indian judiciary has approached the concept of bail and undertrial detention. Over time, in response to systemic delays, overcrowding, and the disproportionate effect of incarceration on socio-economically disadvantaged individuals, the Supreme Court has developed principles that favor liberty, justice, and fair procedure. This section aims to highlight the important judicial pronouncements and developments in undertrial prisoners’ rights, particularly in relation to bail, and how they fit into the broader discourse on justice and legal access.

A) Hussainara Khatoon v. State of Bihar:

In this landmark case, the Supreme Court incorporated the right to a speedy trial as an essential part of Article 21. This case revealed the shocking picture of undertrial prisoners who are kept in confinement for a long period without either the commencement or conclusion of a trial, even exceeding the maximum sentence prescribed for the alleged offense. The Supreme Court held that such a delay was a violation of a person’s liberty and emphasized the importance of delivering justice in a speedy manner and providing legal aid to those undertrial prisoners. This marked a paradigm shift in the approach, considering the rights of undertrial prisoners and highlighting the inequalities faced by those belonging to a socio-economically weaker section of society[7]

B) . Gudikanti Narasimhulu v. Public Prosecutor:

In this landmark case, the Supreme Court laid down a foundational principle stating, “bail is the rule and jail is an exception,” and emphasized that a person’s liberty ought not to be infringed unless compelling reasons exist. This highlights the importance of considering the nature of the offense and whether the accused might flee from justice while granting or refusing bail, and not a punitive approach, as seen in many cases. Importantly, this judgment also emphasized that pretrial detention ought not to be a form of punishment for those who are presumed innocent.[8]

C) . Moti Ram v. The State of Madhya Pradesh:

The Supreme Court critically analyzed the inequality that exists while granting bail, especially while imposing sureties that are unfavorable to the accused, especially those belonging to the weaker section of society. The  Court observed that while granting bail, it is required to ensure that the bail is reasonable and that the accused is not forced to languish in jail only because he is poor. This judgment is important because, through this, the Supreme Court has pointed out the barriers that exist for the poor to get justice.[9]

D). Maneka Gandhi v. Union of India:: The Supreme Court, while dealing with this issue, expanded the scope of Article 21, which required that any procedure for the deprivation of personal liberty was to be "just, fair, and reasonable." Through this judgment, the Supreme Court moved a step closer to a substantive approach to due process. While dealing with bail, this judgment would require that any restrictive procedure that exists for the accused, especially while dealing with modern technology, would need to be critically analyzed for its reasonability.[10]

E). Satender Kumar Antil v. CBI:

In this modern landmark judgment, the Supreme Court reiterated these classic principles and provided guidelines to ensure that unnecessary arrests are avoided and bail is disposed of. The Supreme Court observed that "the courts must be sensitive to the plight of undertrials and that a separate 'Bail Act' may be formulated to bring uniformity to the system.[11]

Taken together, the above judicial rulings establish a liberty-centric bail jurisprudence that strives to strike a balance between individual rights and the overarching interests of justice, while also highlighting existing structural inequalities in the system.

3.2 The Statutory Shift: From CrPC to BNSS 2023

The above judicial principles have been re-established with renewed significance under the Bharatiya Nagarik Suraksha Sanhita, 2023. In particular, Section 479 of the BNSS[12], which replaces Section 436A of the CrPC, provides that undertrial prisoners serving sentences should be released if they have served one-half of the maximum term of imprisonment that could be imposed for their offence. Although the BNSS attempts to enhance procedural efficiency, including through technological integration, its effectiveness needs to be examined critically in light of the standard of what is just, fair, and reasonable, as outlined in Maneka Gandhi v. Union of India, particularly in situations where digital exclusion may be a barrier for undertrial prisoners.

4.  Digitalization of the Bail System

4.1 E-Courts Phase III

The E-Courts Project, as conceptualized under the National Policy and Action Plan for Implementation of Information and Communication Technology (ICT) in Indian Judiciary, 2005[13], represents a revolutionary step towards the digitalization and modernization of the Indian judicial system. The recently approved Phase III of the E-Courts Project, applicable from 2023 onwards, aims to build a digital, paperless, and integrated judicial system through the development of a common platform for interacting with courts, litigants, and other relevant stakeholders.

In terms of features, the E-Courts Phase III has introduced a range of innovative features in the domain of bail administration in India.

1. For instance, the universalization of e-filing systems allows for the electronic filing of bail applications, thus reducing the use of physical documentation

2. Similarly, the expansion of video conferencing facilities allows for virtual hearings in bail matters, ensuring greater ease and accessibility. Further, through the FASTER (Fast and Secure Transmission of Electronic Records) system [14], introduced by the Supreme Court, digital transmission of bail orders to prison authorities has been made possible, eliminating delays associated with postal mail during the post-bail period.

3. The e-Courts portal and NJDG also enhance transparency through online availability of information, allowing stakeholders to track the status of matters online.

A practical example of this change is evident through the pandemic, where bail hearings for courts across India are being conducted through virtual platforms. Such a change not only ensures the continuity of court proceedings but also highlights the potential for technology to reduce pendency and promote timely decision-making.

4.2 Interoperable Criminal Justice System (ICJS)

The Interoperable Criminal Justice System (ICJS)[15] is an integrated digital system that aims to allow real-time data exchange between major components of the criminal justice system, including law enforcement agencies, courts, prisons, and forensic centers. The role of the Interoperable Criminal Justice System in the decision-making process during bail applications is critical, allowing decision-makers to access detailed information about the accused, including first information reports, criminal history, and charge sheets. The use of such data minimizes delays in the decision-making process, thereby speeding up the adjudication process.

However, there are certain risks associated with data-driven decision-making, including a major risk associated with the use of ICJS, which involves the possibility of bias in decision-making. The availability of previous data would likely be a major factor in decision-making, thereby increasing the chances of denying bail to people from certain communities, given that such communities already have a higher representation in criminal data. The use of ICJS would thus necessitate certain checks and balances to ensure that technological advancements do not compromise the principles of justice.

4.3 E-Prisons Project

The E-Prisons Project[16], undertaken by the National Crime Records Bureau (NCRB), is a centralized database of all the imprisoned individuals, capturing data related to the admissions, the status, and the duration of the imprisonment. In the context of bail, the E-Prisons Project is instrumental in identifying the undertrial prisoners eligible to be released under the “half-sentence” provision, now codified under Section 479 of the BNSS, 2023.

However, the potential of the E-Prisons Project is being undermined by the large gap between the implementation and the actual application. Even though the data related to the undertrial prisoners is easily accessible, often there is no mechanism or system to take administrative action or to send alerts to the Legal Aid Defense Counsel (LADC) when the prisoner is eligible to be released on bail. The fact is that the imprisoned individuals are not being released, even after they have become eligible to be released, and this reiterates the point that the mere availability of data does not automatically translate into administrative action.

4.4 Limitations of Digital Bail Infrastructure

Although digitalization has marked a tremendous leap forward, there are certain limitations and hindrances in its proper implementation and functioning.

1. The first and foremost limitation is the digital divide, which has been a problem in our country for a long time and has resulted in a gap in terms of internet accessibility and digital literacy levels among citizens.

2.In addition, the digitalization process from video conferencing has also created a communication paradox, whereby undertrial prisoners are not able to communicate effectively and privately with their lawyers from a jail cell, thereby violating the Right to Effective Counsel. Similarly, security and privacy issues are also a concern, and they remain a critical problem in this regard.

5.1 The Connectivity Divide: Geographies of Exclusion

The digitalization of the Indian judiciary under the e-Courts Project, with a special focus on Phase III, represents a paradigm shift in the technology-driven administration of justice in India. Thousands of courts are being brought online, and the significant public investments made towards the conduct of virtual courts and paperless systems of justice indicate a structurally transformed system of justice administration in India. However, the achievement of this digitalization remains a far cry in the real world of Indian justice administration. While the Supreme Court and some of the High Courts are functioning with the aid of highly digitalized systems of justice administration, the district courts and prisons, where the majority of the undertrial prisoners interact with the system, are struggling with significant infrastructural challenges like the lack of stable internet connectivity, inadequate hardware facilities, and irregularities in the electricity supply. This dichotomy has resulted in a fragmented digital system of justice administration where access to justice is now contingent upon the territorial location of the citizen.[17]

One of the significant consequences of this digital divide in the Indian system of justice administration is the creation of what may be called the ‘digital delay thesis.’ While the digitalization of the Indian justice system is expected to result in the accelerated administration of justice, the technological challenges at the trial level are resulting in the adjournment of cases. While video-conferencing failures are common at the trial level, the lack of publicly available statistics on video-conferencing failures obscures the problem further and results in a lack of transparency and accountability in the system of justice administration in India.

The infrastructural gap identified in the present context leads to the emergence of disparities in the administration of justice. Metropolitan courts, with the support of better connectivity and allocation, are enabled to conduct hearings more effectively, while the same is not true in the case of district courts and penal facilities, where the process is often marred by interruptions and an absence of procedural fairness. The emergence of a ‘network adjournment’ can be conceived as a hypothetical scenario wherein an undertrial prisoner, being handicapped from being produced effectively before the court owing to the lack of connectivity, is presented as merely an absent ‘presence,’ or ‘black screen,’ as if merely a pixelated ‘presence’ on the screen. The emergence of such scenarios speaks to the phenomenon of dehumanized adjudication and the potential erosion of the right to a fair hearing as guaranteed under Article 21.

In the present context, the emergence of the connectivity divide speaks to the proposition that the mere development of digital infrastructure is not enough to ensure the equitable administration of justice. While the country has registered significant success in the development of technological capacity, the failure to ensure the implementation of the same at the grassroots level is likely to result in the perpetuation, as opposed to the mitigation, of disparities.

5.2 Digital Literacy: Barriers to Cognitive Access

The move towards a digitally enabled judiciary is often seen through the prism of physical access, including the availability of technology, internet connectivity, and digital infrastructure. However, for the vast majority of India's undertrial prisoners, the actual challenge is that of cognitive access, or the ability to comprehend and effectively engage with digital legal systems. Digital justice is premised on a level of technological literacy that is at odds with the socio-economic circumstances of prison populations, a significant percentage of whom hail from disadvantaged backgrounds and lack formal educational qualifications.

From a theoretical perspective, true access to justice under Article 21 entails more than just the availability of digital interfaces; rather, it entails what may be termed cognitive justice, or the ability to derive meaning from legal processes and information. For instance, digital interfaces such as NJDG and e-filing often display information in English and use legal and technical jargon, effectively denying those individuals who do not possess the requisite linguistic and digital competency to understand such terminology and jargon. Thus, in such a scenario, a digitally accessible bail order would be no different from one that is not, and hence, procedural justice entails not just access, but also comprehension[18], and in the absence thereof, transparency remains partial and incomplete.

Further, such linguistic and technological barriers within digital interfaces result in what may be termed “double exclusion,” whereby undertrials are unable to understand digital updates and information due to linguistic limitations and, at the same time, do not possess the requisite technological competency to effectively utilize digital interfaces. This effectively results in a tremendous loss of agency, whereby the accused party comes to be largely reliant on lawyers, clerks, and even prison authorities to even obtain rudimentary information regarding his or her case. Unlike physical court hearings, digital hearings result in a loss of agency whereby the accused party comes to be a mere participant, effectively transferring control from the individual to the operator of the device.

In this context, the concept of digital poverty can be conceived as a structural legal disadvantage, particularly when the international norms are moving in the direction of placing a greater emphasis on the concept of technological parity in the legal arena. As the State is mandating the use of digital modes in bail hearings and adjudications, it is, in effect, creating an obligation to ensure the provision of technological inclusivity, including the provision of digital legal aid support. In the absence of the same, the transparency in the digital courts would continue to remain an illusion, which would not be in line with the constitutional requirements of just, fair, and reasonable procedures.

5.3 Intermediary Exploitation: The Emergence of the ‘Digital Middlemen' While the digitalization of the Indian judiciary, exemplified by the e-Courts Project and the Interoperable Criminal Justice System (ICJS), sought to promote greater transparency and reduce the pre-existing ‘middleman’ syndrome in the Indian judicial courts, the reality is that this goal has been achieved to only a limited extent. While the digitalization of bail status, orders, and court cases sought to promote greater interface between individuals and the judiciary, the lack of digital literacy and access has meant that the digitalization of the judiciary has merely resulted in the ‘re-emergence of the ‘middleman’ in more opaque and less regulated forms.

This transformation is enabled by the emergence of a new form of information asymmetry. While the data is technically “publicly available” through the NJDG, it is functionally inaccessible to the undertrial prisoner and his/her families because they lack the requisite device, literacy, and/or technological competence. This leads to the emergence of the “visibility gap,” wherein a variety of intermediaries, ranging from the court clerk to the informal agent and the prison contact, assume the role of digital gatekeepers and often charge for services that are ostensibly free, including checking the bail status, downloading/printing orders, and facilitating communication with legal counsel. The very idea of transparency is thus undermined.[19]

In addition, the classic forms of procedural corruption have given way to what may be characterized as “digital mediation corruption.” Unlike the earlier forms, wherein the misfiling of documents was part of the deliberate strategy, the contemporary forms of exploitation are enabled by unverifiable digital claims, ranging from server failures to portal errors and/or technical delays. The inability of the undertrial prisoner to verify the outputs of the system enables the exploitation by the mediator, and the payment is made with little risk of detection. The problem is further complicated in the rural and semi-urban areas by the emergence of the “cyber café judiciary,” wherein private digital service providers function as access points to the legal system, despite lacking legal training and/or oversight.

In structural terms, the move indicates an extension of the reconfiguration of power within the justice system. In other words, digitalization has not necessarily led to the democratization of access in an unqualified sense, but rather an extension of control in the operational sphere to those who possess technological capital. Insofar as access to the justice system is predicated on the ability to interact with the digital platform, the intermediary becomes a key actor, mediating the relationship between the individual and the state in the first instance. As a result, rather than creating a direct and transparent interface, the digital system potentially creates a new form of exclusion, which challenges the very notion of the just and accessible state.

5.4 Exploratory Insights: Indicative Findings from Stakeholder Interactions

To overcome the divide between the law as analyzed and the law as lived, the present analysis is augmented by the inclusion of exploratory insights drawn from informal interactions with advocates and legal aid workers, as well as an observational engagement with the bail process in the district jurisdictions. Although these are merely indicative and not exhaustive, they add an important critical dimension to the analysis of the digital divide in the criminal justice system. The analysis shows that the move to digital processes was not an even transition, especially in the domain where undertrial prisoners engage most immediately with the system.

One important finding from such interactions has been the observation that digital hearings tend to be formalities, and not necessarily avenues for advocacy. This has been a finding from a variety of interactions, with legal practitioners stating that the “human element” in legal representation has been diluted in digital hearings, where there has been a lack of engagement and communication between the legal practitioner, client, and judge in real time. The inability to communicate in private with undertrial detainees, and the prevalence of technological glitches such as audio delays and disconnections, has meant that digital hearings tend to be mechanical exercises, and in many cases, such technological glitches have become normalized, leading to “routine adjournments,” or what can be termed “network adjournments,” similar to traditional adjournments in a digital form

Another interesting finding was the persistence of a profound awareness gap, where the main stakeholders who would benefit from digital transparency—undertrial prisoners—remain unaware of existing digital platforms like NJDG. Despite State claims of data accessibility, undertrial prisoners lack the means to access information regarding the status of their cases. There is a cycle of dependence on intermediaries, thus reinforcing an existing pattern of digital gatekeeping. In this respect, it can be argued that there is an “information divide,” where existing socio-economic inequalities are further exacerbated.

6. Algorithmic Bias & AI Risks:

Algorithmic bias is defined as “systematic discrimination, mostly undesired, embedded in an automated system due to flawed data, design, or machine learning processes.” [20][21]One good example of this is predictive policing tools that unfairly discriminate against marginalized communities or hiring tools that give preference to some demographics over others due to biases in the training data. While human biases have the privilege to be challenged in court, algorithmic discrimination has no such privilege and is based on mathematical obscurities.

6.1 Predictive tools:

Predictive tools utilizes machine learning to predict where crimes will likely occur or who is likely to commit them. The police department provides historical crime data to this software, and it produces a risk score or a heat map. This all makes sense, right? If you can predict where a robbery is likely to occur on a Tuesday night, you can have police waiting for it. The Chicago Police Department and Los Angeles Police Department began utilizing this software, such as PredPol, now called Geolitica, in the early 2010s to reduce crime rates through more efficient resource allocation.

However, here is where things go wrong with this whole concept. The historical crime data does not reflect where crime occurs. It simply reflects where police have historically been deployed. The over-policed black and Latino neighborhoods have more arrest records, not because crime occurs more frequently within them, but because more police are deployed within them. If this data is utilized within an algorithm, it creates a cycle where more police are deployed to over-policed neighborhoods, leading to more arrests within them. It is a cycle where systemic bias is laundered through data.

6.2 Navigating Black Box:

There are various subfields within AI, and one of the major ones is Machine Learning (ML), which is concerned with the development of models that are able to make predictions or decisions based on a given data set. Deep Learning (DL) is a subset of ML that is based on a conceptual model of the human brain, utilizing artificial neural networks arranged in multiple hidden layers, also known as Deep Neural Networks (DNNs). Thus, DL systems are ML systems, but the converse is not true.

The predictive process starts with an input layer, which represents the raw data input into the system, followed by many layers of "hidden layers," in which each "neuron" is designed to represent a mathematical function that transforms the input data into an output, creating the final prediction. The word "deep" refers to the high number of these "hidden layers." Significantly, the more layers the system has, the more complex and opaque the decision-making process becomes, so that it is impossible for a human to comprehend the process by which the final prediction is made.[22]

This process results in the "black box effect," which is characterized by the lack of transparency and interpretability of the internal processes of the system. While the "black box effect" occurs in ML, in DL, the complexity of the DNNs increases the effect, which can lead to the violation of the most fundamental legal principles, including the principles of due process and equality before the law, when these algorithms are increasingly being deployed in high-risk areas, including recruitment processes or the assessment of creditworthiness.

So, there is a fundamental incompatibility between the accuracy of the system and the transparency of the system, since the high degree of accuracy in the predictions made by the system, which is due to the high dimensionality of the mathematical optimisation process, far exceeds the cognitive capabilities of the human mind, creating an incompatibility between the accuracy of the system and the ability of the human mind to comprehend the process, which results in the opacity of the system, so that as the accuracy increases, the transparency decreases, leaving the individual in a state of defencelessness under the law.

6.3 Global Case Studies:

 6.3 (1) COMPAS:

COMPAS is a risk assessment tool developed by Northpointe, now known as Equivant, to assist courts in making informed decisions about pre-trial release, sentencing, and parole. The COMPAS algorithm make an assessment on the risk of recidivism based on an offender’s history using an algorithm developed to evaluate an offender's risk of recidivism based on factors that include criminal history, substance abuse, and social-familial risk. In May 2016, ProPublica, an investigative media organization, published an expolizive article titled ‘Machine Biea’, which accused the COMPAS algorithm of bias against African-American defendants. According to ProPublica, the COMPAS algorithm is almost twice as likely to wrongly label African-American defendants as future criminals as white defendants. In response to ProPublica’s claims. The technical report claimed that the COMPAS algorithm meets the criteria for statistical fairness, with emphasis on equal positive and negative predictive values.

In the Loomis v. Wisconsin case, Eric Loomis, who was sentenced using the COMPAS algorithm, challenged the use of the algorithm in the sentencing of defendants[23]. Loomis argued that the use of COMPAS in the determination of sentences is a) in violation of due process and b) based on gender assessments. The appellant argued that he had been subjected to discrimination on the grounds of gender, being male. The Supreme Court of Wisconsin ruled on the Loomis case in July 2016, upholding the use of the COMPAS algorithm in the determination of sentences. They also noted the potential biases of the algorithm, citing gender bias and racial bias. It mandated that judges be informed of the limitations of the COMPAS algorithm, noting that it has the potential to discriminate based on race and racial bias.

6.3.2 Algorithmic Tenant Screening Systems, USA:

SafeRent Solutions, among other companies, has been sued on claims of its algorithm systematically gives lower ratings to Black and Hispanic applicants compared to white applicants.48 The algorithm has also been accused of discriminating against those with a voucher by not considering its value.

6.3.3 Dutch Child Benefits Scandal, Netherlands:

While it is not an AI system used in the courts, the scandal concerned an algorithm used by the Dutch tax authorities that wrongly accused thousands of families, mostly of immigrant background, of child benefit fraud. The case illustrates how discriminatory algorithms can result in serious injustice.

6.3.4 UK Post Office Horizon Scandal, UK:

This was another case where an IT system, whose defects led to the wrongful prosecution of various post office workers for theft and false accounting.51 It is a timely reminder of the danger of relying on computer systems without proper precautions and human checks.

6.3.5 Facial Recognition algorithms:

One of the most important areas of concern in AI bias relates to facial recognition software, as they pose one of the biggest risks to human rights. This is due to a lack of diversity in the data sets used, leading to a much higher error rate for darker-skinned individuals, especially women.

In a test carried out by the ACLU on Amazon’s Rekognition algorithm, 28 members of Congress were incorrectly matched to a mugshot database. Almost 40% of these errors were from members of Congress who were people of colour,[24] even though they comprised only 20% of Congress.[22] This has already been seen to play out in real life, such as when Robert Williams, a Black man, was wrongly arrested in Detroit and kept in jail for 30 hours after a facial recognition algorithm wrongly matched his photo from his driver’s license with a mugshot database from a shoplifting suspect.

7. Critical Analysis: The paradox of digital justice:

The digitalisation of justice systems is, in essence, a paradox. While digitalisation through technologies like online courts, electronic filing systems, video conferencing, and AI tools is expected to usher in unprecedented efficiency, cost savings, and increased access to justice , these technologies are at the same time expected to create new barriers for the most vulnerable members of society, effectively turning what was once seen as a bridge into a gatekeeping tool . This critical analysis seeks to examine who really stands to gain from digital justice systems, who continues to be locked out, and whether digitalisation is working for or against the fundamental principles of justice. As scholars put it, "digital justice means fairness, inclusion, and ethics in every stage of digital development." It asks us to think not just about who has access to connectivity, but who designs the technology, whose interests it serves, and who benefits from its use?

7.1 Who Benefits from Digital Justice?

The primary beneficiaries of digital justice systems are those individuals who have what researchers have called the "capability, representation, and inclusion" to effectively interact in digital legal systems. This includes:

a) Digitally Literate Individuals: Those with high digital literacy skills can conduct legal transactions speedily, access information on their own, and use the online court portal without assistance. Research has also shown that those with high digital literacy skills can harness digital tools for learning and research, while others are left behind, unable to participate fully.

b) Socio-Economically Advantaged Users: Individuals with stable internet connections, updated technology, and flexible schedules are greatly benefiting from remote hearings and online filing. One user of the tribunal stated that she and her family were in a privileged position in that we've usually got broadband, we've got computers and all the rest of it.

C) Institutional Beneficiaries: Apart from the users of justice services, the justice institutions themselves derive significant benefits from digitalization. The institutions enjoy "cost savings, efficiency gains, and expanded services." The administrative burdens of the institutions may be reduced, the processing of court cases may be done in less time, and the infrastructure costs may be reduced. But as this analysis will show, efficiency in justice institutions does not necessarily mean equitable access.

7.2 Who is Excluded from Digital Justice?

The digital exclusion in the context of justice is not just a nuisance; it is a matter of justice. As the digitalization of judicial processes increases, what previously required a physical visit to a court or administrative office may now require a stable internet connection, a functioning device, and the capacity to read the interface.

As per research, several populations still face the risk of exclusion due to:

a) Poor Connectivity: This is in spite of the general assumption that everyone is connected. It is being found that 13% adults still do not use the internet, which is the same since 2014, with one in three adults never using a computer to go online. The disparity in the world is even more alarming. Even for people who use the internet regularly, they face the issue of poor internet connection. It is being noticed that only few companies provide a stable internet connection in India that too only in urban areas, the people of rural area still face exclusion from digital justice due to poor connectivity.

b) Digitally Unskilled: Lack of access but with competence gives rise to the "second-level digital divide," which refers to inequality in skills and not in access. This means that people spend long hours online but still lack the digital literacy to evaluate sources, use research tools, or manage data privacy. According to one study, digital access without digital literacy creates the illusion of inclusion.

c) Persons with Disabilities: Another issue is that digital justice systems are not accessible for everyone. Users who need sign language interpretation will face technological challenges, while users with visual or motor disabilities will find the technology impossible to use.

 d) Homelessness or Precarious Housing: Perhaps the most disconcerting aspect of this phenomenon is the fact that "those without a home can be excluded from the digital proceedings altogether. Without reliable access to the internet or the necessary technology, they were virtually silenced in digital proceedings.

8. Conclusion

Digitization of the bail process in India is an essential milestone in improving efficiency and ensuring that individuals are entitled to speedy trials, enshrined under Article 21 of the Indian Constitution. Nevertheless, one key point emphasized in this research is the inherent contradiction between digitization and democracy in the context of the judicial process. Although such measures attempt to improve accessibility, they can perpetuate socio-economic disparities. Indeed, access to digital technologies, illiteracy, and lack of assistance make such efforts more disadvantageous to undertrial prisoners with a marginalized socio-economic background.

Hence, digital justice in its current state is not synonymous with substantive justice. In fact, digitalization reveals yet another digital gap in the criminal justice system. Technology does not exist in a vacuum; instead, it can reflect various forms of discrimination and injustice that already exist in society. Therefore, failure to address such contradictions might cause digital justice to become a type of digital exclusion.

Recommendations

a) Implementation of a Hybrid Bail System: A combination system comprising both online and offline components needs to be put in place. Undertrial prisoners need to be provided an alternative of using the offline method for filing bail applications and attending bail hearing processes so as to guarantee them justice irrespective of their technological incapacity.

2. Setting up of Digital Legal Aid Clinics: Digital legal aid clinics need to be established in district courts, prisons, and rural areas. This is because there will be a necessity of having legal aid lawyers and para-legal volunteers who will guide the undertrials on how to utilize digital platforms.

3. Design of Inclusive and Accessible E-Court Platforms: E-court platforms, as well as other digital systems, need to be redesigned in such a way that they become easily navigable, multi-lingual, and mobile-friendly. Incorporation of voice-guided assistance could enhance access by illiterate individuals.

4. Closing the Digital Divide

There is a need for investments aimed at improving digital infrastructure in rural and poor areas. Besides, digital training should be organized to enable judges, lawyers, prison officials, and undertrial prisoners use digital platforms effectively.

5. Monitoring and Regulation of Algorithmic and AI-based Instruments: Careful consideration needs to be taken when introducing algorithms and AI tools into bail decisions. Ensuring transparency and the presence of a mechanism for human control is crucial to avoid the risks of discrimination and to ensure that judicial discretion is not undermined by any algorithmic black box.

6. Enhancing Accountability Measures: Periodic audits of digital justice solutions and the publication of relevant data will be indispensable. This would assist in identifying and addressing inequalities while ensuring that technological innovation complies with the Constitution.

9. References

 Cases

1. Hussainara Khatoon v. State of Bihar, A.I.R. 1979 S.C. 1369 (India).

2. Gudikanti Narasimhulu v. Public Prosecutor, (1978) 1 S.C.C. 240 (India).

3. Moti Ram v. State of Madhya Pradesh, A.I.R. 1978 S.C. 1594 (India).

4. Maneka Gandhi v. Union of India, A.I.R. 1978 S.C. 597 (India).

5. Satender Kumar Antil v. Central Bureau of Investigation, (2022) 10 S.C.C. 51 (India).

6.State v. Loomis, 881 N.W.2d 749 (Wis. 2016).

B. Legislation

1. INDIA CONST. art. 21.

2. Bharatiya Nagarik Suraksha Sanhita, No. 46 of 2023, India Code.

3. Code of Criminal Procedure, No. 2 of 1974 (India) (repealed).

C. Government Reports & Policy Documents

1.Nat’l Crime Records Bureau, Prison Statistics India 2023 (Ministry of Home Affairs 2024).

2. Supreme Court E-Committee, Vision Document for Phase III of e-Courts Project (2023).

3. Ministry of Home Affairs, Interoperable Criminal Justice System (ICJS) Report (2022).

4. Nat’l Crime Records Bureau, e-Prisons Project Report (2021).

5.Law Comm’n of India, 268th Report on Bail Reforms (2017).

D. Books & Articles

1. Upendra Baxi, The Crisis of the Indian Legal System (1982).

2. Marc Galanter, Law and Society in Modern India (1989).

3. Sourav Das, Access to Justice and the Digital Divide in India, 12 Indian J.L. & Tech. 45 (2020).

4. Algorithmic Justice: Regulating Bias in AI-Driven Judicial Decision-Making in India: Nickeled And Dimed (2025)

E. International & Comparative Sources

1. Julia Angwin et al., Machine Bias, ProPublica (May 23, 2016), https://www.propublica.org.

2. United Nations Development Programme, Digital Strategy 2022–2025 (2022).

3. World Bank, World Development Report 2016: Digital Dividends (2016).

4. American Civil Liberties Union, Amazon Rekognition Test Report (2018).

5. Navigating the black box: AI bias and the future of the burden of proof in the EU: UNIO-EU Law Journal (2025)

F. Websites

E-Courts Mission Mode Project, https://ecourts.gov.in.

National Judicial Data Grid, https://njdg.ecourts.gov.in.

[1] 1. National Crime Records Bureau, Prison Statistics India 2023 (Ministry of Home Affairs).

[2] 2. INDIA CONST. art. 21.

[3] 3. Hussainara Khatoon v. State of Bihar, (1980) 1 SCC 81.

[4] 4. Bharatiya Nagarik Suraksha Sanhita, No. 46 of 2023.

[5]  E-Committee, Supreme Court of India, eCourts Project Phase III Vision Document (2023).

[6] . Jan van Dijk, The Deepening Divide: Inequality in the Information Society (Sage Publications 2005).

[7] Hussainara Khatoon v. State of Bihar, (1980) 1 SCC 81.

[8]. Moti Ram v. State of Madhya Pradesh, (1978) 4 SCC 47.

11. Maneka Gandhi v. Union of India, (1978) 1 SCC 248.

[11] 12. Satender Kumar Antil v. CBI, (2022) 10 SCC 51.

[12]  Bharatiya Nagarik Suraksha Sanhita, 2023, § 479.

[13] Department of Justice, National Policy and Action Plan for Implementation of ICT in Indian Judiciary (2005).

[14]  In Re: FASTER (Fast and Secure Transmission of Electronic Records), Suo Motu Writ (C) No. 5 of 2020 (Supreme Court of India).

[15] Ministry of Home Affairs, Interoperable Criminal Justice System (ICJS) Project Report (2022).

[16] National Crime Records Bureau, e-Prisons Project Documentation (2023).

[17] . E-Committee, Supreme Court of India, eCourts Phase III Vision Document (2023).

[18] . Amartya Sen, Development as Freedom (Oxford Univ. Press 1999).

[19]  Transparency International, Corruption in Judicial Systems Report (2020).

[20] Algorithmic Justice: Regulating Bias in AI-Driven Judicial Decision-Making in India: Nickeled And Dimed (2025)

[21] Julia Angwin et al., Machine Bias, ProPublica (May 23, 2016), https://www.propublica.org.

[22] Navigating the black box: AI bias and the future of the burden of proof in the EU: UNIO-EU Law Journal (2025)

[23] State v. Loomis, 881 N.W.2d 749 (Wis. 2016)

[24] American Civil Liberties Union, Amazon Rekognition Test Report (2018).