A Jurisprudential Inquiry into Algorithmic Justice in Heinous Crime Cases in India

INTRODUCTION:

In contemporary India, where heinous crimes attract the gravest punishments, ensuring procedural fairness and judicial transparency is of paramount importance. The judiciary contends with burgeoning caseloads, prolonged delays, and inconsistencies in sentencing, particularly in cases involving life imprisonment or capital punishment. To address these challenges, courts are increasingly exploring technological tools, including AI, to assist in bail and sentencing decisions. Algorithmic justice refers to the use of computational models and predictive analytics to provide data-driven insights that may enhance efficiency, consistency, and impartiality in judicial outcomes. International experiences, such as the ^[1]Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) risk-assessment tool in the United States, illustrate both the potential benefits and limitations of such systems. In India, initiatives like ^[2]the Supreme Court Portal for Assistance in Court Efficiency demonstrate a cautious engagement (SUPACE) with AI in managing complex legal information but ‘^[3]the Black Box issue which makes AI inherently opaque and antithetical to the principles of due process that due process demands.’ In this context, one must ask: even with the promise of artificial intelligence, can human judges retain ultimate authority to ensure justice, fairness, and constitutional integrity in the adjudication of heinous crimes?

Heinous Crime And Indian Criminal Jurisprudence

Heinous crimes in India constitute offenses that attract the gravest penal sanctions, including life imprisonment or the death penalty, encompassing murder, aggravated rape, terrorism-related offenses, and acts of waging war against the state. Under the ^[4]Bhartiya Nyaya Sanhita these offenses are codified with stringent provisions, which covers murder and associated acts, while terrorism-related crimes are addressed comprehensively to safeguard societal security. These legislative measures reflect the state’s intent to deter the most egregious violations while embedding procedural safeguards that uphold the “rule of law”. The Supreme Court of India has reiterated that capital punishment is the exception rather than the norm. In ^[5]Bachan Singh v. State of Punjab (1980) 2 SCC 684, the Court articulated the ‘rarest of rare’ doctrine, reserving the death penalty for cases demonstrating extreme moral culpability. In ^[6]Machhi Singh v. State of Punjab (1983) 3 SCC 470, it emphasized the imperative of carefully evaluating aggravating and mitigating circumstances before imposing the maximum penalty. Collectively, these precedents underscore that sentencing in heinous crimes must remain proportionate, judicious, and consonant with constitutional guarantees, particularly ^[7]Articles 14 and 21, which ensure equality before law and the right to life and personal liberty.

Given the complexity and high stakes inherent in adjudicating such offenses, there is growing interest in integrating AI to assist judicial discretion in risk assessment, bail determinations, and sentencing recommendations. While tools such as the Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) illustrate AI’s potential, ultimate responsibility must reside with human judges to safeguard due process, fairness, transparency, and accountability. ^[8]“The use of AI in criminal law adjudication and sentencing presents a dynamic shift in the paradigm from the traditional model of justice, based on the application of criminal justice norms in accordance with the subjective human satisfaction of a human judge, to a more robust and systematic one, which is based on data, statistics, permutations and calculations by an AI to reach deterministic conclusions.”

 

How AI is Changing India’s Justice System

Artificial intelligence (AI) has increasingly permeated the criminal justice system in India, especially in adjudicating heinous crime cases. As Agarwal remarks, ^[9]“Many algorithms used in digital markets operate behind the scenes, making it difficult to understand how they are making decisions and to identify instances of algorithmic collusion”. This algorithmic opacity introduces novel challenges for ensuring fairness and transparency in legal determinations.

AI-based tools, particularly those used in judicial case management, have shifted approaches to justice delivery. As emphasised by Roy, ^[10]“In India, the courts have envisaged an AI system that can sift through thousands of pages of criminal petitions and extract relevant information out of them and present a summarized version to the judge, revolutionizing the approach to case management”. Such interventions potentially streamline case handling but risk reducing the human dimension in complex decision making. Moreover, AI-powered research platforms are changing professional legal practice. ^[11]“AI-powered legal research platforms like Manupatra and SCC Online have transformed the way legal professionals in India conduct research. These platforms use AI algorithms to analyse and categorize vast amounts of legal data, making it easier for lawyers and judges to find relevant case law and statutes”. Yet, questions of bias and data integrity persist.

Empirical research points to systemic dangers: ^[12]“Key findings include underrepresentation of marginalised communities in AI datasets, lack of ethical AI guidelines, risks of surveillance misuse and limited awareness of AI ethics among developers and policymakers. A novel insight is the danger of addressing AI bias reactively, as this could entrench further harm”. Jurisprudential inquiry into algorithmic justice demands vigilance, ensuring that efficiency does not undermine equity, due process, and human agency.

 

Can AI Improve Fairness in Justice?

Artificial intelligence offers significant opportunities to promote fairness in India’s criminal justice system. As observed, ^[13]“AI in Indian courts is transforming justice delivery by enabling judges to access relevant case law and legal provisions quickly, thus reducing delay and human error.” This procedural efficiency underscores AI’s promise to enhance equitable access to justice, particularly for marginalized parties burdened by systemic delays. Moreover, AI technologies can standardize judicial processes to mitigate inconsistent sentencing patterns, a long-standing jurisprudential concern. Yet, as Sharma cautions, ^[14]“concerns about bias, fairness, and transparency persist in AI adoption, necessitating robust safeguards and human oversight.” Indeed, the Kerala High Court has established a vital precedent by proclaiming that ^[15]“under no circumstances AI tools are used as a substitute for decision-making or legal reasoning.” This insistence on human primacy ensures that AI augments rather than replaces judicial discretion, protecting the fairness inherent in due process. Further, AI’s capacity to bridge linguistic and accessibility barriers is crucial in a diverse country like India, strengthening the substantive equality principle by democratizing legal information. However, the realization of AI’s fairness potential demands ethical governance frameworks that institutionalize transparency, auditability, and accountability at every stage of AI implementation. Hence, while AI proves a powerful instrument for advancing fairness in criminal justice, its efficacy depends critically on embedding it within a jurisprudential framework that safeguards fundamental rights and guarantees meaningful human control.

Challenges and Ethical Concerns

The incorporation of artificial intelligence (AI) into India’s criminal justice system, while promising transformative efficacy, simultaneously presents profound challenges and ethical quandaries that strike at the heart of jurisprudence and constitutional guarantees. Paramount among these challenges is the inherent risk of algorithmic bias and systemic discrimination perpetuated through opaque AI frameworks. As Sharma critically observes, ^[16]“concerns about bias, fairness, and transparency persist in AI adoption, necessitating robust safeguards and human oversight.” Such biases are particularly pernicious in India’s diverse socio-legal context where disadvantaged groups risk magnified marginalization through AI-driven decision-making, undermining the constitutional mandate of equality before law. The jurisprudential dilemma of accountability emerges acutely given AI’s autonomous decision-making capacity. Indian criminal law’s foundational components—mens rea and actus reus—face complexities in attribution of liability where AI operates beyond direct human control or exhibits unpredictable behavior. Sayyed elucidates that ^[17]“the attribution of criminal liability to AI systems raises fundamental questions about intent, accountability, and the ethical responsibility of human actors,” underscoring the lacuna in existing frameworks to prosecute or regulate AI conduct under established penal doctrine. This ontological challenge calls for doctrinal evolution to delineate responsibilities among AI developers, deployers, and state actors. Ethical concerns further intensify surrounding the intrusion upon privacy rights and the erosion of transparency in the adjudicatory process. The ‘black box’ nature of many AI algorithms thwarts meaningful judicial scrutiny and judicial reasoning; contravening principles of due process elaborated in Indian constitutional jurisprudence. The ability to audit and explain algorithmic logic remains rudimentary, impeding effective challenges to erroneously adverse or prejudiced AI outputs. The Kerala High Court’s recent directive that ^[18]“AI tools are not to substitute decision-making or legal reasoning” reflects judicial recognition of the imperative for human oversight to safeguard procedural fairness and ethical integrity. Moreover, the proliferation of AI-generated deepfakes and synthetic evidence strikes at evidentiary sanctity vital to fair trial guarantees. As noted by commentators on emerging AI jurisprudence, deepfakes ^[19]“have the potential to manipulate evidence, tarnish reputations, and spread misinformation, significantly impacting the integrity of legal proceedings.” This technological threat mandatorily demands forensic innovation and judicial guidelines to authenticate digital evidence, preventing miscarriages of justice catalysed by fabricated realities. The ethical design and deployment of AI also foreground profound questions about social justice. AI systems risk entrenching existing disparities through discriminatory data sets and result in disproportionate targeting of vulnerable communities, mandating a jurisprudential commitment to recalibrate AI governance towards accountability, transparency, and human rights compatibility. This is consonant with principles advocated by global human rights norms, which stress equitable technological governance. In sum, integrating AI into India’s criminal justice system necessitates a multi-dimensional ethical and legal framework embedding transparency, procedural fairness, individual rights, and rigorous human oversight. Courts must remain vigilant guardians against AI tools that could erode justice’s humane core, mandating adaptive legislation and doctrinal innovation to meet 21st-century challenges. Without such resolute jurisprudential commitment, AI’s promise to enhance justice risks becoming an instrument of injustice itself.

Case Study

The integration of Artificial Intelligence (AI) within India’s criminal justice system has steadily progressed, yielding transformative applications alongside profound jurisprudential challenges.

1. SUPACE: Revolutionizing Judicial Research and Adjudication

 

The Supreme Court Portal for Assistance in Court Efficiency (SUPACE), launched in April 2021, represents the Supreme Court’s landmark adaptation of AI to judicial administration. Designed in collaboration with IIT Madras, SUPACE “assists judges by extracting relevant case laws, legal provisions, and factual information from voluminous filings, expediting case adjudication without replacing judicial discretion”. During its pilot in criminal proceedings at the Bombay and Delhi High Courts, SUPACE demonstrated its capacity to reduce judicial research burdens significantly, directly addressing the backlog of over 44 million pending cases by ^[20]“improving the quality and timeliness of judicial decisions while preserving human judgment”. SUPACE’s AI tools ^[21]“provide informational assistance, but judicial independence remains inviolable,” underscoring the necessity to safeguard fairness and due process in the AI augmentation of justice. ^[22]Nevertheless, legal scholars caution that reliance on historic data demands stringent audits to preclude propagation of systemic biases embedded in past jurisprudence.

 

2. CMAPS and Predictive Policing by Telangana Police

 

The Telangana Police’s Crime Mapping, Analytics, and Predictive System (CMAPS) exemplifies AI’s operational impact on law enforcement through predictive analytics harnessing historical crime data, geo-spatial insights, and socio-economic indicators. ^[23]CMAPS aims to forecast crime hotspots and enable resource optimization, complemented by AI-powered facial recognition technologies facilitating real-time suspect identification. While praised for enhancing policing efficiency and strategic deployment, this initiative raises profound ethical and constitutional concerns. Critics argue these tools ^[24]“engender privacy violations and carry severe risks of discriminatory profiling,” disproportionately affecting marginalized communities, thereby risking violation of civil liberties and procedural justice. This case illuminates the tension between AI-enabled public safety measures and fundamental rights guarantees, pressing the need for robust data protection legislation and judicial oversight.

 

3. AI in Bail, Parole, and Sentencing

 

A Double-Edged Sword Several Indian states have introduced AI-assisted risk assessment tools to assist judicial officers in bail, parole, and sentencing decisions. These systems synthesize offender histories, flight risk evaluations, and socio-demographic data to generate predictive scores intended to guide magistrates in delivering consistent, objective rulings. However, Manupatra case commentaries and SCC Online reports caution that these AI algorithms embody ^[25]“the ‘black box’ phenomenon, with opaque decision-making processes that threaten the transparency and accountability essential to fair adjudication”. Furthermore, ^[26]there exists an acute risk that unexamined algorithmic outputs will exacerbate existing social and judicial inequities by automating biases inherent in the datasets. The jurisprudential message is clear: AI in sentencing requires rigorous transparency mandates, human supervision, and mechanisms for appellate review to ensure the AI serves as an aid, not an arbiter, of justice.

 

Collectively, these case studies demonstrate India’s ambitious yet cautious engagement with AI in criminal justice, highlighting the need to rigorously balance technological innovation with enduring principles of fairness, equality, and rights protection

 

Recommendations

 

The integration of Artificial Intelligence (AI) in India’s criminal justice system avails transformative prospects to alleviate delays, enhance decision-making quality, and improve law enforcement efficacy. Yet, realizing these benefits requires a principled approach prioritizing fairness, transparency, privacy, and human rights. The following recommendations emerge from scholarly and policy analyses addressing the technological, ethical, and legal dimensions of AI deployment in criminal justice.

 

1. Establish a Comprehensive and Dedicated Legal Framework:

 

India must enact specific, granular legislation governing AI applications in criminal justice with clear provisions on accountability, bias mitigation, data protection, and procedural fairness. Existing legislation, such as the Information Technology Act, though useful, is insufficient to address AI’s complex challenges. As noted by scholars, ^[27]“India’s AI governance framework remains nascent and inadequate to regulate the nuanced implications of AI in judicial and policing contexts”. ^[28]A specialized regulatory authority with oversight powers over AI technologies, able to enforce audits and penalties, is indispensable to prevent unregulated AI-enabled abuses while encouraging responsible innovation.

 

2. Mandate Algorithmic Transparency and Explainability:

 

All AI systems must be designed with inherent transparency and interpretability to uphold constitutional guarantees of fair trial and reasoned decision-making. ^[29]The “black box” nature of certain AI tools obstructs meaningful judicial and defendant scrutiny, risking miscarriages of justice. ^[30]Legislators and the judiciary should mandate that AI algorithms deployed for judicial assistance, sentencing, sentencing prediction, or policing decisions fully disclose their operational logic and data sources, enabling challenge and review. This transparency will also foster public trust in increasingly AI-mediated justice processes.

 

3. Ensure Data Integrity, Security, and Diversity:

 

The quality, diversity, and protection of datasets feeding AI systems are critical. Biased or incomplete data can compound systemic discrimination. As highlighted in recent policy reports, ^[31]“Representational inequities in training data propagate algorithmic bias, disproportionately disadvantaging marginalized groups”. ^[32]India’s emerging data protection framework must align tightly with AI requirements to guarantee anonymization, user consent, limited data retention, and enforced security protocols. Additionally, AI must be regularly tested for fairness and accuracy across diverse demographic profiles.

 

4. Embed Robust Human Oversight and Judicial Education

 

AI must complement rather than supplant human discretion and judgment in criminal justice. “Every AI-generated recommendation should be subject to critical human evaluation, especially when liberty or constitutional rights are concerned.” ^[33]Judicial officers require continuous education on AI’s mechanics, benefits, and limitations to make informed evaluations of AI input. Such capacity building forms the bedrock of accountability and rights protection in AI-augmented adjudication.

 

5. Foster Multi-Stakeholder and Inclusive Governance

AI governance must incorporate diverse voices from legal experts, technologists, civil society, and especially affected communities. This inclusive approach ensures AI tools are socially contextualized and ethically designed. “Engaging marginalized groups prevents further marginalization and fosters trustworthy AI systems.” Additionally, public transparency initiatives must educate citizens on AI’s role and safeguards within criminal justice, empowering democratic oversight.

6. Promote Iterative Piloting, Evaluation, and Adaptation

 

AI deployment should initially proceed through rigorously monitored pilots with continuous impact assessment. ^[34]These evaluations must measure fairness, accuracy, and unintended social consequences to iteratively refine AI systems. ^[35]Regional contextualization is key given India’s socio-legal diversity, requiring adaptable models rather than uniform national prescriptions

Conclusion

 

The integration of artificial intelligence into India’s criminal justice system holds transformative promise to streamline judicial processes, enhance law enforcement efficiency, and address the persistent backlog that undermines timely justice. Systems like SUPACE and AI-driven predictive policing exemplify how technology can assist accurate decision-making while optimizing resource allocation. However, this promise must be tempered with caution, as AI’s opacity, potential for bias, and privacy concerns pose fundamental threats to fairness, equality, and constitutional rights. As Sharma correctly stresses, ^[36]“while AI can enhance crime prediction, detection, and offender management, it also raises important legal and ethical concerns”. Given these challenges, India must adopt a comprehensive legal and regulatory framework that mandates transparency, accountability, human oversight, and data integrity. Judicial training and multi-stakeholder engagement are essential to ensure that AI remains a decision-support tool—not a replacement for human judgment—thus upholding democratic values and due process. As emphasized in official government reports, ^[37]“integrating AI in India’s judiciary and law enforcement must be guided by principles of fairness, inclusivity, and rights protection”. By harmonizing innovation with ethics and rigorous oversight, India can leverage AI’s potential to create a criminal justice system that is not only more efficient but more just and equitable for all.

[1] Northpointe Inc, Correctional Offender Management Profiling for Alternative Sanctions (COMPAS Risk Assessment Tool, 2013)

[2] Supreme Court of India, ‘Supreme Court Portal for Assistance in Court Efficiency (SUPACE)’ (2021)

[3] Raajdweep Vardhan^* and Naveen Kuma, “Automated Adjudication and Criminal Law Sentencing: An Analysis from the Lens of Due Process and Human Rights”(2024), 151 SCC online CNLU LJ (11)

[4] Bhartiya Nyaya Sanhita, 2023.

[5] Bachan Singh v. State of Punjab (1980) 2 SCC 684.

[6] Machhi Singh v. State of Punjab (1983) 3 SCC 470.

[7] Indian Constitution, Article 14 and 21.

[8] “Raajdweep Vardhan^* and Naveen Kuma, “Automated Adjudication and Criminal Law Sentencing: An Analysis from the Lens of Due Process and Human Rights”(2024), 151 SCC online CNLU LJ (11).

[9] Agarwal, S, ‘Changing Dynamics of Algorithmic Collusion: An Analytical Study’ (SCC Online, 2023).

[10] Roy, P, ‘Algorithmic Decision-making and Criminal Trials in India’ (JSTOR, 2022).

[11] Choudhary, D & Jain, R, ‘Impact of Artificial Intelligence on Judicial System’ (Manupatra, 2020).

[12] David, D et al., ‘Algorithmic Bias and Discrimination in India: A Looming Crisis’ (HeinOnline/SAGE Journals, 2025).

[13] Bhat Y, ‘AI in Indian Courts – Transforming Justice Delivery’ (Legasis Blog, 2025)

[14] Sharma A, ‘Artificial Intelligence in the Indian Criminal Justice System’ (SDGS Review, 2025)

[15] Gandhi A and Thakur R, ‘Kerala High Court’s New AI Guidelines Set National Standard for Judicial Integrity’ (SSRana Blog, 2025)

[16] Sharma A, ‘Artificial Intelligence in the Indian Criminal Justice System’ (SDGS Review, 2025)

[17] Sayyed H, ‘Artificial Intelligence and Criminal Liability in India’ (Tandfonline, 2024)

[18] Gandhi A and Thakur R, ‘Kerala High Court’s New AI Guidelines Set National Standard for Judicial Integrity’ (SSRana Blog, 2025)

[19] Kehl D et al, ‘Algorithms in the Criminal Justice System: Assessing Use of AI’ (JSTOR, 2019)

[20] PIB, ‘Use of Artificial Intelligence in Supreme Court’ (Government of India, 2025)

[21] SCC Online, ‘SUPACE and the Digital Transformation of Judicial Research in India’ (2024).

[22] CLPR Blog, ‘Artificial Intelligence and Judicial Bias’ (Centre for Law and Policy Research, 2021)

[23] Manupatra, ‘AI-Powered Predictive Policing in Telangana’ (2023).

[24] Legasis, ‘AI in Indian Courts – Transforming Justice Delivery’ (2025)

[25] SCC Online, ‘AI in Bail and Parole: Enhancing Fairness or Risking Opaqueness?’ (2025).

[26] Manupatra, ‘Risks and Governance of AI-Assisted Sentencing’ (2025).

[27] Sharma A, Sharma S C, Soni S D et al, ‘Artificial Intelligence in the Indian Criminal Justice System: Advancements, Challenges, and Ethical Implications’ (SDGS Review, 2025)

[28] PIB, ‘Integrating AI in India’s Judiciary and Law Enforcement’ (Government of India, 2025)

[29] Sayyed H, ‘Artificial Intelligence and Criminal Liability in India’ (Tandfonline, 2024)

[30] SCC Online, ‘AI in Bail and Parole: Enhancing Fairness or Risking Opaqueness?’ (2025).

[31] Manupatra, ‘AI-Powered Predictive Policing in Telangana’ (2023).

[32] CLPR, ‘Artificial Intelligence and Judicial Bias’ (Centre for Law and Policy Research, 2021)

[33] SCC Online, ‘SUPACE and the Digital Transformation of Judicial Research in India’ (2024).

[34] NLS Repository, ‘The Perils and Promises of Artificial Intelligence in Criminal Justice’ (2025).

[35] DAKSH Network, ‘Algorithmic Accountability in the Judiciary: A Rights-Based Approach’ (2024).

[36] Sharma A, ‘Artificial Intelligence in the Indian Criminal Justice System’ (SDGS Review, 2025).”

[37] PIB, ‘Integrating AI in India’s Judiciary and Law Enforcement’ (Press Information Bureau, 2025).

 

BY- Dhwani Amin, 2nd year
Student of Gujarat National Law University, Silvassa Campus