International Journal of Law, Politics and Governance
DOI: 10.64823/ijlpg.2601007
Criminal investigations traditionally relied on tangible evidence such as fingerprints, bloodstains, documents, eyewitness testimony, and physical objects, whose material nature enabled relatively straightforward verification. However, the rapid expansion of digital technologies has transformed the evidentiary landscape. Mobile phones, CCTV systems, cloud platforms, and social media now generate huge amounts of electronically stored information that frequently serve as crucial evidence in criminal proceedings. Reflecting this shift, the National Institute of Standards and Technology (NIST) defines digital evidence as electronic data possessing evidentiary value in legal processes, underscoring its growing importance within contemporary criminal justice systems.[1]
The shift from physical evidence to digital evidence has brought undeniable advantages. Investigators can reconstruct timelines with unprecedented precision, trace communications across networks, and analyse large volumes of information in ways that would have been unimaginable a generation ago. The successful prosecution of several major criminal and terrorism cases has demonstrated the evidentiary value of digital records, including internet logs, surveillance footage, geolocation data, and electronic communications.[2]
The digital environment that generates evidence also facilitates its manipulation. Images can be altered, audio recordings synthetically produced, and videos fabricated with remarkable realism. Consequently, questions of authenticity have become central to evidentiary assessment. Rather than simply determining whether an event occurred, investigators and courts must first establish whether a digital record is genuine and reliable.
The growing reliance on digital evidence has coincided with what may be described as a crisis of authenticity. Advances in artificial intelligence have significantly lowered the barriers to producing highly convincing synthetic content. Deepfakes, generated through machine-learning techniques capable of replicating human appearance, voice, and behaviour, have moved from experimental demonstrations into practical tools capable of influencing public discourse, commercial transactions, and legal proceedings.[3]
The challenge extends beyond detecting fabricated material. The emergence of sophisticated deepfakes has generated what Chesney and Citron describe as the “liar’s dividend,” whereby authentic evidence may be dismissed as manipulated because convincing falsification has become technologically feasible. Consequently, evidentiary assessment must establish both the falsity of fabricated records and the authenticity of genuine ones. Recent forensic research, including studies conducted by the National Institute of Standards and Technology (NIST), indicates that deepfake detection systems often experience reduced effectiveness when applied to newly developed generative AI models, highlighting persistent difficulties in maintaining reliable authentication mechanisms.[4] Advances in generative AI have outpaced traditional evidentiary safeguards. Digitally manipulated content can now be disseminated widely before verification occurs, creating challenges for courts and investigators and exposing limitations in existing legal frameworks governing authenticity, reliability, and evidentiary assessment.
Criminal justice institutions increasingly employ artificial intelligence to analyse digital evidence, automate document review, support forensic examinations, and identify patterns in large datasets. Proponents contend that these systems enhance investigative efficiency and enable the processing of information at scales beyond human capacity.[5]
Blockchain technology has been proposed as a tool for enhancing evidentiary integrity through tamper-evident and cryptographically secured records. Such systems may strengthen chain-of-custody documentation and transparency in evidence handling, although concerns regarding scalability, governance, and legal admissibility remain unresolved.[6]
This paper examines the changing nature of evidentiary reliability in digital criminal justice systems. While existing scholarship often treats artificial intelligence, blockchain, deepfakes, and predictive policing separately, limited attention has been given to how these technologies collectively affect the reliability, admissibility, and legitimacy of digital evidence. Accordingly, the principal research question guiding this paper is: How do artificial intelligence, deepfakes, blockchain authentication mechanisms, and predictive policing technologies affect the reliability, admissibility, and legitimacy of digital evidence in criminal justice proceedings?
Two leading subsidiary questions are also examined: Can blockchain-based authentication mechanisms effectively safeguard evidentiary integrity against digital manipulation? And are existing legal and regulatory frameworks adequately equipped to address risks associated with AI-generated and algorithmically mediated evidence?
This paper proceeds from the hypothesis that technological innovation alone does not guarantee evidentiary reliability. Rather, reliability depends on the interaction of technical verification mechanisms, institutional safeguards, legal oversight, forensic expertise, and procedural accountability. The analysis therefore focuses on the conditions under which emerging technologies strengthen or weaken the reliability, admissibility, and legitimacy of digital evidence.
The analysis focuses on the United States, the United Kingdom, and India, highlighting differing regulatory responses to common technological challenges. Particular attention is given to the Bharatiya Sakshya Adhiniyam, 2023, which expands recognition of electronic evidence while raising new questions concerning authenticity, admissibility, and evidentiary value.[7]
The paper proceeds in six stages. It begins with a review of the literature and identification of the research gap, followed by the theoretical and methodological framework. Subsequent sections examine artificial intelligence, deepfakes, blockchain-based authentication, and predictive policing. The paper concludes by evaluating comparative legal responses and proposing measures to strengthen the reliability, admissibility, and legitimacy of digital evidence.
Table 1.1: Technology-Evidence Nexus in Criminal Justice
Technology | Primary Function | Principal Benefit | Principal Risk |
Artificial Intelligence | Analysis and Decision Support. | Investigative Efficiency and Pattern Recognition. | Opacity and Explainability Challenges. |
Deepfakes | Synthetic Content Generation. | Creative and Educational Applications. | Manipulation and Evidentiary Uncertainty. |
Blockchain | Verification and Record Preservation. | Integrity and Traceability. | Input Reliability and Governance Concerns. |
Predictive Policing | Crime Forecasting and Resource Allocation. | Strategic Deployment of Resources. | Algorithmic Bias and Over-Policing. |
Source: Compiled by the Authors from Ronald J. Allen & Alex Stein, “Evidence, Probability, and the Burden of Proof”, (2013) 55(3) Arizona Law Review 557-602; Karen Yeung, “Algorithmic Regulation: A Critical Interrogation”, (2018) 12(4) Regulation & Governance 505-523; Andrew Guthrie Ferguson, The Rise of Big Data Policing: Surveillance, Race, and the Future of Law Enforcement, New York University Press, New York, 2017; Yisroel Mirsky & Wenke Lee, “The Creation and Detection of Deepfakes: A Survey”, (2021) 54(1) ACM Computing Surveys 1-41; and European Union Blockchain Observatory and Forum, Blockchain for Government and Public Services, 2022.
The increasing digitization of social, economic, and governmental activities has transformed the evidentiary foundations of criminal justice systems, with investigations increasingly relying on electronic communications, surveillance footage, metadata, cloud-stored information, and social media records. Early research in digital forensics focused on forensic acquisition, preservation, and chain-of-custody procedures. Studies by Casey (2011) and NIST guidelines on digital forensics emphasised maintaining evidentiary integrity during collection and analysis, highlighting the susceptibility of electronic records to alteration, duplication, and deletion, thereby creating distinct challenges for authenticity and reliability.[8] As digital technologies became embedded in everyday life, scholarship expanded beyond forensic procedures to address privacy, surveillance, accountability, and procedural fairness. A recurring concern is the tension between evidentiary abundance and reliability. Although contemporary investigations generate enormous amounts of digital information, greater data availability does not necessarily facilitate accurate fact-finding. Instead, courts and investigators must assess the authenticity, reliability, and verification of digital records. As Kuner argues, digital technologies have transformed information into both an evidentiary resource and a source of legal uncertainty.[9]
Artificial intelligence has become one of the most extensively discussed technological developments within contemporary criminal justice scholarship. Broadly understood, AI refers to computational systems capable of performing tasks that traditionally required human cognition, including pattern recognition, classification, prediction, and decision support.[10] Within criminal justice systems, artificial intelligence is used in facial recognition, forensic analysis, predictive policing, and risk assessment. Proponents argue that AI enhances investigative efficiency by analysing large datasets, identifying suspicious patterns, detecting cyber threats, and assisting digital forensics. Studies also indicate that AI-assisted tools can reduce investigative backlogs and accelerate evidence analysis.[11]
At the same time, a growing body of scholarship questions assumptions regarding algorithmic neutrality. Scholars such as Cathy O'Neil, Ruha Benjamin, and Virginia Eubanks argue that algorithmic systems frequently inherit biases embedded within historical datasets and institutional practices.[12] Scholars argue that seemingly objective algorithmic systems may reproduce existing social inequalities, particularly in criminal justice contexts where they influence investigative priorities and risk assessments. Concerns also arise from algorithmic opacity, as many AI systems operate through processes that are difficult to interpret. This “black box” problem challenges principles of transparency and accountability, raising questions about defendants’ rights to understand algorithmically informed decisions. While some scholars support greater disclosure, others warn that extensive transparency may compromise proprietary interests and security.[13] These competing concerns reflect a broader debate regarding the role of AI in legal and evidentiary processes. Although AI has been widely studied as a decision-support tool, its simultaneous capacity to enhance evidentiary verification and create evidentiary uncertainty, particularly through the emergence of deepfakes, remains comparatively underexplored.
Few technological developments have generated as much concern regarding evidentiary authenticity as deepfakes. Produced through advanced machine-learning techniques, deepfakes enable the creation of highly realistic synthetic audio, video, and image content that may be difficult to distinguish from authentic recordings.[14]
Initial academic discussions often treated deepfakes as a problem of misinformation and political communication. Researchers highlighted the potential use of synthetic media in election meddling, disinformation campaigns, and under the belt attack on character and reputation. Furthermore, documented incidents involving manipulated political speeches and fabricated public statements attracted widespread scholarly and policy makers attention.[15]
More recent scholarship has extended the debate into the legal domain. Chesney and Citron argue that deepfakes pose distinctive challenges because they undermine the evidentiary value traditionally associated with visual and audio recordings.[16]
Historically, photographs and videos were often regarded as highly persuasive forms of proof. Deepfake technology complicates that assumption by making realistic fabrication increasingly accessible. A particularly influential concept is the “liar’s dividend,” whereby individuals may dismiss authentic recordings as manipulated. The threat, therefore, extends beyond fabricated material to the erosion of trust in genuine digital records. Scholars argue that deepfakes endanger not only factual accuracy but also the epistemic foundations of legal institutions. At the same time, some caution against technological determinism, noting that advances in deepfake detection and digital forensic techniques continue to enhance verification capabilities. Consequently, evidentiary reliability is increasingly understood as the product of an ongoing contest between fabrication and authentication rather than the inevitable triumph of deception.[17]
While deepfake scholarship focuses largely on evidentiary manipulation, blockchain scholarship has generally concentrated on evidentiary verification. Blockchain technology emerged from discussions surrounding cryptocurrencies but has since attracted broader interest as a mechanism for establishing trust in digital environments.[18]
A blockchain functions as a distributed ledger in which transactions are recorded across multiple nodes and protected through cryptographic mechanisms. Because entries become difficult to alter retrospectively, scholars have suggested that blockchain systems may strengthen chain-of-custody management and evidence preservation.[19]
Several studies have explored blockchain-based evidence management systems within law enforcement and judicial contexts. Researchers argue that immutable audit trails may improve transparency and reduce opportunities for tampering. Pilot projects in jurisdictions including China, Estonia, and selected European Union initiatives have demonstrated the potential value of blockchain-assisted evidentiary systems.[20]
Yet enthusiasm has not been universal. Critics observe that blockchain can verify the integrity of stored information but cannot independently guarantee the accuracy of information entered into the system. The familiar principle of "garbage in, garbage out" remains relevant. If inaccurate or manipulated evidence is recorded at the outset, blockchain technology may merely preserve the error with greater efficiency.[21]This limitation highlights an important distinction between evidentiary integrity and evidentiary authenticity. Whereas integrity concerns whether evidence has been altered after collection, authenticity concerns whether it accurately reflects the event or circumstance it purports to represent. Blockchain systems are particularly effective in safeguarding integrity through immutable audit trails, but they cannot by themselves establish authenticity. The distinction becomes especially significant when blockchain technologies are used alongside AI-generated or synthetic media, where the reliability of the original content may itself be contested.
The emergence of predictive policing has generated one of the most vigorous debates within contemporary criminal justice scholarship. Predictive systems employ statistical models and machine-learning techniques to forecast crime patterns, identify potential crime hotspots, or assess risk profiles.[22] Supporters of these initiatives contend that predictive policing allows law-enforcement agencies to deploy limited resources more efficiently. Programmes such as PredPol and Chicago's Strategic Subject List were introduced partly on the premise that data-driven forecasting could improve crime prevention and operational effectiveness.[23] Critics, however, question whether predictive systems merely automate existing biases. Historical crime data often reflects earlier policing practices, reporting patterns, and institutional priorities. If these datasets contain embedded inequalities, predictive systems may reproduce them under the appearance of scientific objectivity.[24]Scholarship on predictive policing therefore remains divided between claims of improved resource allocation and concerns regarding bias, transparency, accountability, and civil liberties.
Three major debates emerge in the literature. First, scholars disagree on whether technological systems enhance or undermine evidentiary reliability, with competing emphasis on efficiency versus manipulation and bias. Second, opinions differ regarding the balance between technological innovation and procedural safeguards; and third, scholars remain divided over whether existing legal frameworks can adequately accommodate emerging technologies or require fundamentally new regulatory approaches
Although scholarship on artificial intelligence, deepfakes, blockchain, and predictive policing has expanded, these technologies are typically examined in isolation. Comparatively little attention has been given to their interaction within a shared evidentiary framework. This gap is significant because criminal justice institutions often deploy these technologies simultaneously. Accordingly, this paper develops an integrated framework linking evidentiary reliability, algorithmic governance, and digital trust, examining the conditions under which technological systems strengthen or weaken the reliability, admissibility, and legitimacy of digital evidence.
The convergence of artificial intelligence, blockchain technologies, and predictive analytics has transformed contemporary approaches to digital evidence. Understanding this transformation requires attention to the legal and technological developments that have shaped the present evidentiary landscape.
Timeline 1: Evolution of Digital Evidence and AI Governance (2008-2024)
Year | Development | Significance |
2008 | Satoshi Nakamoto publishes Bitcoin: A Peer-to-Peer Electronic Cash System. | Introduces blockchain technology and distributed trust mechanisms. |
2014 | Anvar P.V. v. P.K. Basheer | Establishes stricter requirements for the admissibility of electronic evidence in India. |
2019 | Growth of deepfake technologies and major regulatory debates. | Raises concerns regarding authenticity and digital manipulation. |
2020 | Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal. | Clarifies the legal position on Section 65B certification. |
2022 | Europol publishes Facing Reality? Law Enforcement and the Challenge of Deepfakes. | Highlights criminal justice implications of synthetic media. |
2023 | Bharatiya Sakshya Adhiniyam was enacted. | Modernizes Indian evidentiary law for the digital age. |
2023 | UK AI White Paper published. | Introduces a pro-innovation approach to AI regulation. |
2024 | Regulation (EU) 2024/1689 (EU AI Act) was adopted. | Establishes the world's first comprehensive AI regulatory framework. |
Source: Compiled by the Author from Satoshi Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System” (2008); Anvar P.V. v. P.K. Basheer, (2014) 10 SCC 473; Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal, (2020) 7 SCC 1; Europol, Facing Reality? Law Enforcement and the Challenge of Deepfakes, 2022; Bharatiya Sakshya Adhiniyam, 2023; Department for Science, Innovation and Technology, A Pro-Innovation Approach to AI Regulation, HM Government, 2023; and Regulation (EU) 2024/1689 (Artificial Intelligence Act).
At the heart of every criminal justice system lies a deceptively simple question: can the evidence be trusted? Whether a court is evaluating a witness statement, a forensic report, or a digital recording, the legitimacy of judicial decision-making ultimately depends upon the reliability of the information presented. Evidentiary reliability theory provides a useful framework for examining this issue because it focuses on the conditions under which evidence can be regarded as sufficiently trustworthy to support legal conclusions.[25]
Traditionally, evidentiary reliability has been assessed through authentication, chain-of-custody procedures, corroboration, expert testimony, and judicial scrutiny. The emergence of artificial intelligence and synthetic media complicates these mechanisms, as fabricated images, audio, and video may appear authentic while genuine records may be challenged as false. Consequently, reliability can no longer be evaluated solely through traditional safeguards. This study employs evidentiary reliability theory to examine three interrelated questions: how digital evidence can be authenticated, how evidentiary integrity can be preserved, and what institutional safeguards are required to maintain confidence in technologically mediated evidence
While evidentiary reliability focuses on the trustworthiness of information, algorithmic governance examines the growing role of computational systems in shaping institutional decision-making. The concept refers broadly to the use of algorithms, automated processes, and data-driven systems to guide, influence, or structure governance activities.[26]
The growing use of artificial intelligence in criminal justice reflects a broader shift toward algorithmic governance. Predictive policing, facial recognition, risk assessment, and AI-assisted forensic tools are increasingly employed to manage large volumes of information and improve efficiency. However, critics challenge assumptions of algorithmic objectivity, noting that historical datasets often embed institutional biases and social inequalities. Consequently, machine-learning systems may reproduce existing patterns of discrimination while appearing technologically neutral.[27]
The issue is not merely technical. It is also legal and normative. Criminal justice systems derive legitimacy from transparency, accountability, and procedural fairness. Many algorithmic systems, however, operate through highly complex computational processes that remain difficult for external observers to understand. Frank Pasquale's influential description of the "black box society" captures this concern.[28] As algorithmic systems increasingly influence decisions affecting individuals, their opaque internal operations often make it difficult to understand, scrutinize, or challenge the basis on which those decisions are made. Accordingly, this study employs algorithmic governance as a framework for analysing how technological systems, institutional contexts, and regulatory structures shape evidentiary assessment and criminal justice decision-making.
Digital trust constitutes a fundamental element of legal systems, underpinning confidence in evidentiary reliability and institutional legitimacy. The growth of interconnected digital environments has complicated traditional indicators of trustworthiness. Consequently, digital trust frameworks emphasise mechanisms such as transparency, traceability, auditability, verification procedures, and institutional oversight to maintain confidence in technologically mediated evidence and decision-making.[29]
Blockchain technologies are often presented as trust-generating mechanisms because they employ cryptographic verification and distributed recordkeeping. However, trust cannot be reduced to technology alone, as blockchain systems cannot independently verify the accuracy of original information. Consequently, trust in criminal justice systems remains dependent on human judgment, institutional safeguards, and legal accountability, making technology a complement rather than a substitute for established oversight mechanisms.
Although evidentiary reliability theory, algorithmic governance, and digital trust emerge from different intellectual traditions, they share a common concern with technology and institutional legitimacy. Together, these perspectives provide an integrated framework for analysing how technological systems influence evidentiary reliability, decision-making processes, and institutional trust within contemporary criminal justice systems.
Table 3.1: Integrated Analytical Framework
Framework | Core Question | Application in this Study |
Evidentiary Reliability Theory | Can the evidence be trusted? | Authentication, Admissibility, and Integrity. |
Algorithmic Governance | How do algorithms shape decisions? | AI systems and Predictive Policing. |
Digital Trust Frameworks | How is trust established in digital environments? | Blockchain, Verification, and Accountability. |
Source: Compiled by the Authors from Ronald J. Allen & Alex Stein, “Evidence, Probability, and the Burden of Proof”, (2013) 55(3) Arizona Law Review 557-602; Karen Yeung, “Algorithmic Regulation: A Critical Interrogation”, (2018) 12(4) Regulation & Governance 505-523; Frank Pasquale, The Black Box Society: The Secret Algorithms That Control Money and Information, Harvard University Press, Cambridge, 2015; and European Union Blockchain Observatory and Forum, Blockchain for Government and Public Services, 2022.
The study adopts an interdisciplinary socio-legal methodology combining doctrinal analysis, comparative research, and qualitative case studies. Comparative analysis focuses on developments in the United States, the United Kingdom, and India, while the empirical component draws upon documented deepfake incidents, blockchain-based evidence management initiatives, predictive policing programmes, and recent Indian developments in digital evidence governance.
The relationship between technology and evidence has continually evolved, with innovations such as fingerprint analysis and DNA profiling transforming forensic identification and criminal investigations. Artificial intelligence appears poised to similarly reshape contemporary evidentiary practices, although its implications remain contested.[30]
Artificial intelligence is commonly understood as the capacity of computational systems to perform tasks that traditionally require human cognition, including pattern recognition, classification, prediction, and decision support. Within criminal justice systems, AI is increasingly employed to process large volumes of digital information, analyse communications data, identify suspicious behavioral patterns, examine surveillance footage, and assist forensic investigations.[31]
Contemporary criminal investigations generate enormous quantities of digital information, including communications, financial records, geolocation data, and video evidence, making AI-assisted analysis increasingly attractive. While AI can process and prioritise information at scales beyond human capacity, efficiency does not necessarily ensure reliability. The same technologies that support evidentiary verification may also introduce new forms of uncertainty. Consequently, as AI becomes more deeply integrated into the production, analysis, and authentication of evidence, concerns regarding transparency, accountability, and reliability have assumed growing importance within criminal justice systems.
Artificial intelligence has become deeply embedded in modern forensic practice. Digital forensic laboratories increasingly utilize machine-learning systems to assist with evidence review, image analysis, cybercrime investigations, and pattern recognition. What previously required months of manual examination can often be completed within days or even hours.[32] AI has demonstrated considerable utility in analysing digital communications. Machine-learning systems can identify relevant information, reconstruct timelines, detect communication networks, and flag anomalies within large datasets. These capabilities have proven particularly valuable in investigations involving organised crime, fraud, cybercrime, and online child exploitation.[33] Moreover, AI-assisted tools are increasingly employed in facial recognition, image analysis, and cybercrime investigations, reflecting their growing importance in addressing complex transnational threats.[34]
Nevertheless, many AI systems operate probabilistically, generating predictions and classifications rather than definitive conclusions. Studies have identified risks of false positives and false negatives, particularly in facial-recognition technologies. The wrongful arrest of Robert Williams in Michigan following an erroneous facial-recognition match illustrates the potential consequences of excessive reliance on algorithmic outputs in criminal investigations.[35] Such incidents do not invalidate AI technologies but highlight the need for meaningful human oversight. While AI can assist evidentiary analysis, critical evaluation by investigators, forensic experts, and courts remains essential to ensure accuracy, accountability, and reliability.
If AI-assisted forensics represents the constructive face of artificial intelligence, deepfakes reveal its disruptive potential. Deepfakes are synthetic media generated through machine-learning techniques capable of creating highly realistic audio, video, and image content. Early deepfake systems relied primarily upon Generative Adversarial Networks (GANs), while newer generative AI models have further enhanced the realism and accessibility of synthetic content.[36]
Advances in generative AI have made the creation of synthetic images, audio, and video increasingly accessible through commercially available tools. This development has significant implications for criminal justice, where visual and audio records have traditionally been regarded as persuasive evidence. Deepfakes can fabricate events, voices, and locations with considerable realism and can be disseminated widely before verification occurs. Their growing prevalence has prompted concerns among legal scholars, forensic practitioners, and policymakers regarding the reliability and evidentiary value of digital records.
Several documented incidents illustrate the growing evidentiary challenges associated with synthetic media. One widely discussed example emerged during the Russia-Ukraine conflict when a manipulated video falsely depicted Ukrainian President Volodymyr Zelenskyy urging Ukrainian forces to surrender. Although the fabrication was quickly identified, the incident demonstrated how synthetic media can be deployed to influence public perceptions during periods of crisis.[37]
Financial fraud has provided another revealing context. In 2024, international media reported cases in which employees authorized significant financial transfers after participating in video conferences featuring AI-generated impersonations of senior executives. In one widely cited incident involving a multinational company in Hong Kong, fraudsters reportedly used deepfake technology to impersonate corporate leadership and facilitate fraudulent transfers amounting to millions of dollars.[38]
Deepfake pornography has raised significant concerns by producing realistic synthetic images and videos depicting an act that never occurred. Beyond privacy breaches and attacks on dignity, these incidents demonstrate the growing capacity of AI systems to fabricate convincing digital records. Consequently, the evidentiary value of digital recordings increasingly depends on verification processes rather than visual appearance alone.
Paradoxically, the most serious threat posed by deepfakes may not be fabricated evidence itself. Chesney and Citron famously introduced the concept of the "liar's dividend" to describe situations in which individuals exploit public awareness of deepfakes to deny authentic evidence.[39] The possibility of convincing synthetic media has created what scholars describe as a “liar’s dividend,” whereby genuine recordings may be dismissed as fabricated. Consequently, deepfakes undermine confidence in both false and authentic evidence. Unlike earlier forms of digital manipulation, their realism increasingly limits the reliability of unaided human observation, creating greater dependence on forensic authentication techniques such as expert testimony, metadata analysis, and cryptographic verification when assessing digital evidence.
The rise of synthetic media has stimulated parallel efforts aimed at detection and regulation. Researchers have developed numerous AI-assisted detection systems designed to identify manipulation through inconsistencies in facial movements, audio patterns, metadata, compression artefacts, and digital fingerprints. NIST has played a leading role in evaluating the effectiveness of such technologies through its ongoing media-forensics research programmes.[40]
Deepfake detection remains an evolving challenge, as rapidly advancing generative models often outpace existing verification tools. Researchers frequently describe this dynamic as an ongoing contest between generation and detection technologies. In response, governments have introduced regulatory measures, including the European Union’s AI Act of 2024, which establishes a risk-based framework and imposes transparency requirements for certain forms of AI-generated content.[41]
In the United States, regulatory responses to deepfakes remain fragmented, with federal and state measures addressing election-related misinformation, identity fraud, and non-consensual synthetic imagery. Several states, have enacted laws restricting the dissemination of deceptive deepfakes intended to influence elections, while other jurisdictions have targeted the creation and distribution of non-consensual sexually explicit synthetic content. India has similarly intensified policy discussions following the leaks of high profile public figures deepfakes, which generated significant concerns regarding misinformation, reputational harm, and electoral integrity. Concurrently, the Bharatiya Sakshya Adhiniyam, 2023 has strengthened the legal significance of electronic evidence by expressly recognising electronic and digital records as documentary evidence, thereby making questions of authenticity, integrity, and reliability increasingly important in criminal adjudication.[42]
The evidence reveals a recurring paradox: artificial intelligence is increasingly used to authenticate, analyse, and verify digital evidence while simultaneously enabling the creation of highly convincing synthetic content. Consequently, technological solutions alone cannot guarantee evidentiary reliability. Contemporary research suggests that effective verification depends upon the interaction of technological safeguards, forensic expertise, legal standards, and institutional oversight. The central challenge for criminal justice systems is therefore not merely the adoption of AI, but the preservation of trust and confidence in increasingly digital evidentiary environments. As the same technological ecosystem that strengthens evidentiary verification also expands opportunities for manipulation, reliability therefore increasingly depends upon robust governance frameworks capable of ensuring transparency, accountability, and independent scrutiny.
Table 4.1AI and Digital Evidence: The Reliability Paradox
AI Application | Evidentiary Benefit | Evidentiary Risk | Criminal Justice Implication |
AI-Assisted Forensic Analysis | Rapid processing of large volumes of digital evidence; improved pattern recognition and investigative efficiency. | False positives, algorithmic errors, and opaque decision-making processes. | Necessitates human oversight and expert validation of AI-generated outputs. |
Facial Recognition Systems | Enhanced identification of suspects, victims, and missing persons. | Misidentification, demographic/ethnic bias, and privacy concerns. | Raises questions regarding reliability, admissibility, and procedural fairness. |
Deepfake Detection Tools | Assists in identifying manipulated audio, video, and image content. | Detection systems may fail against increasingly sophisticated synthetic media. | Creates a continuous technological arms race between detection and manipulation. |
Generative AI and Deepfakes | Potential applications in training, simulation, and forensic reconstruction. | Fabrication of realistic but false evidence; misinformation and fraud. | Challenges traditional assumptions regarding authenticity and evidentiary trust. |
Voice Cloning Technologies | Accessibility tools, forensic reconstruction, and investigative simulation. | Identity theft, financial fraud, impersonation, and evidentiary deception. | Complicates authentication of audio evidence. |
Blockchain-Assisted Verification Systems | Strengthens chain of custody, traceability, and evidentiary integrity. | Cannot independently verify the authenticity of original content. | Supports preservation but does not eliminate the need for forensic verification. |
Predictive Analytics | Improved resource allocation and crime forecasting. | Reinforcement of historical biases and over-policing of certain communities. | Requires transparency, accountability, and regular auditing mechanisms. |
AI-Generated Investigative Insights | Identifies hidden patterns and relationships in complex datasets. | Risk of overreliance on algorithmic recommendations. | Human judgment remains essential for evidentiary assessment. |
If artificial intelligence has complicated questions of authenticity through the emergence of synthetic media, blockchain technology has been presented as a potential response to the growing crisis of trust in digital environments. Advocates frequently portray blockchain as a technological mechanism capable of strengthening transparency, accountability, and evidentiary integrity. Yet such claims warrant careful examination. Like many technological solutions proposed for criminal justice systems, blockchain offers important possibilities while also revealing significant limitations.
The concept of blockchain first gained prominence through the publication of Satoshi Nakamoto's Bitcoin white paper in 2008.[43] Although initially associated with cryptocurrencies, blockchain technology has gradually attracted interest across a range of sectors, including finance, healthcare, supply-chain management, public administration, and law enforcement. At its core, blockchain functions as a distributed ledger that records transactions across multiple network participants. Each transaction is cryptographically linked to previous records, creating a chronological chain that becomes increasingly difficult to alter retrospectively.[44]
For criminal justice institutions, blockchain offers a potential mechanism for strengthening chain-of-custody management and reducing risks associated with human error, administrative failures, and evidentiary tampering. By enabling independent verification of records, blockchain has been proposed as a response to growing concerns over digital manipulation and deepfakes. However, while blockchain may preserve evidentiary integrity, it cannot by itself establish the authenticity of the original information, leaving a critical evidentiary challenge unresolved.
Although blockchain has generated considerable public attention, its underlying principles are relatively straightforward. A blockchain is essentially a distributed database maintained across multiple computers, commonly referred to as nodes. Transactions entered into the system are grouped into blocks and linked through cryptographic hashes that create an immutable sequence of records.[45] For criminal justice systems, the most promising application of blockchain technology concerns chain-of-custody management. Four blockchain characteristics are particularly relevant for evidentiary purposes: immutability, transparency, traceability, and distributed verification. Together, these features have encouraged the use of blockchain systems for document authentication and the preservation of digital evidence.[46]
For criminal justice systems, the most promising application of blockchain technology concerns chain-of-custody management. Its key evidentiary advantages include immutability, transparency, traceability, and distributed verification. These characteristics are particularly valuable because the chain of custody refers to the documented process through which evidence is collected, transferred, analysed, stored, and ultimately presented in court to ensure its integrity. While traditional and digital recordkeeping systems remain vulnerable to error and manipulation, blockchain-based systems create immutable, time-stamped records of evidentiary transactions, providing transparent audit trails and strengthening evidentiary preservation.[47]
Interest in blockchain-assisted evidence management has expanded considerably during the past decade. Estonia’s digital governance infrastructure,[48] China’s Internet Courts[49], and initiatives supported by the European Blockchain Observatory and Forum[50] illustrate growing institutional interest in distributed ledger technologies for record preservation, authentication, and evidentiary verification. Law-enforcement agencies have similarly explored blockchain-assisted evidence preservation and audit trails. Despite these developments, large-scale adoption remains limited owing to concerns regarding governance, interoperability, privacy, and implementation costs.
The principal advantage of blockchain-based evidence management lies in strengthening evidentiary integrity through tamper-evident records. Blockchain systems also enhance transparency by providing auditable histories of digital evidence and improve accountability through permanently recorded transactions. Additionally, distributed verification mechanisms may facilitate cross-jurisdictional cooperation and the validation of evidentiary records across different legal and institutional systems. [51]These advantages explain why blockchain is frequently described as a technology of trust. Yet trust, as legal scholars have repeatedly observed, cannot be reduced entirely to technological design.
Despite considerable enthusiasm, blockchain is not a panacea. Its principal limitation lies in the distinction between evidentiary integrity and authenticity. While blockchain can demonstrate that information has remained unchanged after entry, it cannot verify the accuracy of the original data. Consequently, manipulated or AI-generated content entered into a blockchain system may be preserved with the same reliability as authentic records.
Questions of governance, scalability, cost, and privacy continue to limit the adoption of blockchain-based evidence systems. Differences in institutional control, technical requirements, and resource availability complicate implementation, while criminal investigations often involve sensitive personal data. Accordingly, scholars caution that blockchain should be viewed as a complementary verification mechanism rather than a substitute for procedural safeguards, judicial oversight, and forensic expertise.[52]
Advances in computing power, data analytics, geographic information systems, and machine learning have transformed policing from a predominantly reactive model toward anticipatory strategies. Within this shift, predictive policing emerged as a prominent application of artificial intelligence, using data-driven analysis to support crime forecasting, resource allocation, and preventive interventions.[53]
Predictive policing employs statistical models, machine-learning algorithms, and historical crime data to forecast criminal activity, identify crime hotspots, and allocate policing resources. However, the approach has generated significant debate, as scholars question whether algorithmic systems merely identify crime patterns or reproduce existing social inequalities embedded within historical data.
Predictive policing draws on criminological theories of hotspot policing, routine activity, and risk management, which identify recurring spatial and temporal patterns in criminal activity. Advances in data analytics and machine learning have transformed these insights into predictive tools, enabling law-enforcement agencies to forecast crime risks and allocate resources proactively rather than merely responding to completed offences. As these systems increasingly shape policing priorities, resource allocation, and enforcement decisions, scholars have begun to view them not merely as forecasting technologies but as instruments of algorithmic governance.[54]
Among the most widely discussed predictive policing programmes was PredPol, later rebranded as Geolitica. Developed through collaborations involving criminologists, mathematicians, and law-enforcement practitioners, PredPol sought to forecast locations where property crimes and other offences were most likely to occur.[55]
PredPol used historical crime data to generate daily forecasts identifying small geographic areas at elevated risk of criminal activity, enabling targeted patrol deployment. Early evaluations reported reductions in certain crime categories, leading supporters to argue that algorithmic forecasting could improve resource allocation, policing efficiency, and operational effectiveness.[56] Subsequent scrutiny focused on the limitations of predictive policing and the assumptions embedded within the datasets on which such systems relied. These concerns contributed to broader debates regarding bias, accountability, and the governance of algorithmic decision-making in criminal justice.[57]
By 2020, the Los Angeles Police Department had terminated its predictive policing programmes amid growing concerns regarding effectiveness, transparency, and civil liberties. The decision became a symbolic moment in broader debates concerning algorithmic governance and criminal justice reform.[58] The PredPol experience therefore highlighted that algorithmic tools cannot be evaluated in isolation from the institutional, legal, and social contexts in which they operate.
If PredPol focused primarily on geographic forecasting, Chicago's Strategic Subject List (SSL) represented a different approach. Rather than identifying high-risk locations, the programme used criminal histories, arrest records, and network associations to generate risk scores for individuals considered at elevated risk of involvement in violent crime. While proponents viewed the programme as a tool for targeted intervention and data-driven governance, empirical evaluations produced mixed results, indicating limited predictive accuracy and difficulties in reliably identifying individuals most likely to engage in future violent offences.[59] While proponents viewed the programme as a tool for targeted intervention and data-driven governance, empirical evaluations produced mixed results, indicating limited predictive accuracy and difficulties in reliably identifying individuals most likely to engage in future violent offences.[60]
Transparency concerns further complicated the Strategic Subject List, as many individuals were unaware of their risk classifications or the criteria used to generate them. Civil liberties advocates questioned the compatibility of such systems with principles of due process and traditional evidentiary standards. The programme's eventual discontinuation highlighted broader practical, ethical, and accountability challenges associated with the use of predictive analytics in criminal justice.
Perhaps no issue has generated greater debate within predictive policing scholarship than algorithmic bias. Contrary to popular assumptions, algorithms do not emerge from neutral technological environments. They learn from data generated by human institutions. Historical crime records, arrest statistics, and enforcement patterns often reflect broader social realities, including economic inequalities, demographic disparities, and institutional priorities.[61]
Machine-learning systems trained on historical data may reproduce existing patterns of inequality and bias. Because algorithmic outputs often appear objective and authoritative, decision-makers may place undue confidence in risk scores and statistical forecasts despite their underlying limitations. Scholars such as Cathy O’Neil, Ruha Benjamin, and Sarah Brayne argue that algorithmic systems can convert historical patterns of disadvantage into future predictions, thereby reinforcing existing social disparities.[62]
Some commentators caution against treating algorithmic bias as uniquely problematic, noting that human decision-making is also susceptible to prejudice, inconsistency, and error. The central issue, therefore, is not whether algorithms should be used, but how they should be governed, regulated, and integrated within accountable decision-making frameworks.
The legitimacy of criminal justice systems depends heavily upon transparency and accountability. Individuals affected by governmental decisions generally possess opportunities to challenge evidence, question assumptions, and contest adverse findings. Predictive policing systems complicate these principles because many machine-learning models operate through computational processes that remain difficult for external observers to understand. Even when source code is available, complex interactions among variables may produce outcomes that are not readily interpretable. Frank Pasquale's notion of the "black box society" remains particularly relevant in this context.[63]
From a legal perspective, such algorithmic opacity raises significant questions concerning disclosure, procedural fairness, and oversight. In response, several jurisdictions have emphasised explainability, auditability, and independent review, recognising that responsibility for algorithmically informed decisions ultimately remains with human decision-makers. Maintaining meaningful transparency, however, becomes increasingly challenging as predictive systems grow more sophisticated.
The experience of predictive policing demonstrates that technological outcomes are shaped by institutional contexts, legal frameworks, and oversight mechanisms. It further illustrates that efficiency does not necessarily ensure legitimacy, as concerns regarding fairness, accountability, and civil liberties may persist; for jurisdictions adopting AI-enabled policing, independent auditing and meaningful oversight remain essential. More broadly, predictive policing demonstrates that technological effectiveness does not eliminate the need for legal accountability and democratic governance.
The rapid adoption of artificial intelligence, deepfakes, blockchain authentication systems, and predictive policing technologies has generated significant regulatory challenges for criminal justice systems. This section comparatively examines legal and policy responses in the United States, the United Kingdom, and India, focusing on how each jurisdiction has sought to address questions of authenticity, accountability, and evidentiary reliability.
The United States has been at the forefront of debates concerning artificial intelligence, deepfakes, predictive policing, and digital evidence, although regulatory responses remain fragmented. Digital evidence continues to be governed primarily through the Federal Rules of Evidence, particularly requirements relating to authentication and reliability.[64] The rise of sophisticated synthetic media has complicated traditional evidentiary standards, prompting several states, including California, Texas, and Virginia, to enact legislation addressing deepfakes in electoral processes, non-consensual intimate imagery, and identity fraud.[65] These concerns intensified following incidents such as the AI-generated robocall impersonating Joe Biden during the 2024 New Hampshire primary, which sought to discourage voter participation.
Artificial intelligence has also attracted increasing federal attention. In 2023, the White House issued Executive Order 14110, titled Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, also referred to as "Executive Order on Artificial Intelligence,” emphasising transparency, accountability, and risk management.[66] Although not directed specifically at digital evidence, the Order reflects growing governmental concern with AI governance.
Predictive policing has generated particularly intense scrutiny. Experiences involving PredPol, Chicago's Strategic Subject List, and other algorithmic systems prompted widespread debate concerning bias, transparency, accountability, and civil liberties. Several police departments subsequently reduced or discontinued predictive policing programmes amid concerns regarding their effectiveness and legitimacy.[67]
The American experience illustrates both the strengths and limitations of decentralized governance. While experimentation across jurisdictions has encouraged innovation, the absence of a comprehensive regulatory framework has resulted in uneven oversight and continuing uncertainty regarding the appropriate use of predictive analytics in criminal justice.
The United Kingdom has adopted a somewhat different approach, placing greater emphasis upon institutional oversight, data protection, and human rights considerations. While the United Kingdom has actively explored the use of artificial intelligence within public administration and law enforcement, regulatory discussions have frequently been framed through the lens of proportionality, accountability, and rights protection.[68] Digital evidence in criminal proceedings is governed by a combination of common-law principles, statutory provisions, and procedural guidance. The Police and Criminal Evidence Act 1984 (PACE), the Criminal Procedure Rules, and various evidentiary guidelines continue to provide the primary legal framework for admissibility and authentication.
The United Kingdom has also been at the forefront of debates concerning algorithmic decision-making within public institutions. Reviews conducted by the Information Commissioner's Office (ICO), the Alan Turing Institute, and parliamentary committees have highlighted the need for transparency and explainability in automated systems.[69] Particular attention has focused on facial-recognition technologies. Judicial scrutiny intensified following the landmark decision in R (Bridges) v Chief Constable of South Wales Police, where the Court of Appeal held that aspects of the police deployment of automated facial recognition technology were unlawful because of deficits in governance and safeguards.[70] The judgment underscored the principle that technological capability does not automatically justify deployment without adequate legal protections.
This emphasis on institutional accountability has also shaped the United Kingdom's response to emerging risks associated with deepfakes and synthetic media. Government reports and parliamentary inquiries have increasingly highlighted concerns relating to misinformation, identity fraud, and digital manipulation while seeking to preserve freedom of expression and innovation. Together, these developments reflect a broader regulatory approach that views digital governance not merely as a technological challenge, but as a question of transparency, oversight, and accountability within public institutions.
India occupies a particularly important position within contemporary debates concerning digital evidence and artificial intelligence. As one of the world's largest digital societies, India has experienced rapid growth in internet connectivity, mobile communications, digital payments, and data-driven governance. These developments have generated significant opportunities for law enforcement while simultaneously creating new evidentiary challenges.[71] The most significant recent legal development is the enactment of the Bharatiya Sakshya Adhiniyam, 2023 (BSA), which replaced the Indian Evidence Act, 1872. The BSA reflects a conscious effort to modernise evidentiary law in light of technological transformation. Electronic and digital records now occupy a more prominent position within the statutory framework, reflecting the growing importance of digital evidence in criminal investigations and judicial proceedings.[72] The BSA builds upon earlier judicial developments concerning electronic evidence, including landmark decisions such as Anvar P.V. v. P.K. Basheer and Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal, which emphasised the importance of authentication and certification requirements for electronic records.[73] Together, these developments demonstrate an increasing judicial awareness that digital evidence requires specialised procedural safeguards.
India has also witnessed growing concern regarding deepfakes and synthetic media. Several high-profile incidents involving manipulated videos of public figures, celebrities, and journalists have generated public debate concerning misinformation and digital authenticity, most notably the widely circulated deepfake depicting Rashmika Mandanna in 2023 and subsequently altered videos involving figures such as Aamir Khan and Amit Shah. These incidents intensified discussions concerning platform accountability, content moderation, and the development of technological mechanisms for detecting synthetic media. Artificial intelligence is simultaneously becoming more prominent within policing and governance. Law-enforcement agencies have explored AI-assisted surveillance, facial-recognition systems, predictive analytics, and data-driven investigative tools. Programmes associated with smart policing initiatives illustrate the expanding role of digital technologies within criminal justice administration.[74]
However, important challenges remain. India continues to grapple with questions concerning privacy, algorithmic accountability, institutional capacity, and forensic infrastructure. While technological adoption has accelerated, governance frameworks continue to evolve. The Digital Personal Data Protection Act, 2023, represents an important step toward strengthening data governance, yet comprehensive AI-specific regulation remains under development.[75] India's experience highlights a broader reality confronting many developing democracies: technological innovation often advances more rapidly than the legal and institutional frameworks designed to govern it. The central challenge lies not merely in adopting new technologies, but in ensuring that safeguards for accountability, transparency, and due process develop alongside them.
A comparative assessment of the United States, the United Kingdom, and India reveals a common pattern: technological innovation has advanced more rapidly than the legal and institutional frameworks designed to govern it. Across all three jurisdictions, artificial intelligence, blockchain authentication systems, and predictive policing technologies have enhanced investigative and evidentiary capabilities while simultaneously generating challenges relating to authenticity, transparency, accountability, and procedural fairness. While the United States has relied primarily on fragmented and sector-specific regulatory responses, the United Kingdom has emphasized institutional oversight, explainability, and rights-based governance. India, meanwhile, has pursued legal modernization through the Bharatiya Sakshya Adhiniyam, 2023, alongside broader efforts to strengthen digital governance and regulatory capacity.
Despite these differences, several common themes emerge. All three jurisdictions increasingly recognize the need to adapt traditional evidentiary frameworks to address AI-generated content, deepfakes, and algorithmically mediated decision-making. They also acknowledge that technologies cannot be treated as neutral or self-validating mechanisms and therefore require meaningful human oversight, transparency, and institutional safeguards.
A central finding of this paper is that evidentiary reliability in the digital age is shaped by the interaction between technological systems and governance frameworks. Although emerging technologies can strengthen verification, improve evidentiary management, and enhance investigative efficiency, their effectiveness ultimately depends upon accountability, forensic expertise, legal oversight, and procedural fairness. The comparative evidence therefore suggests that the most effective regulatory approaches are those that balance technological innovation with democratic values, ensuring that efficiency gains do not come at the expense of fairness, legitimacy, and the rule of law.
Table 7.1: Comparative Regulatory Responses to Emerging Technologies
Issue | United States | United Kingdom | India |
Deepfakes | State-level legislation and federal initiatives. | Regulatory reviews and parliamentary oversight. | Emerging regulatory responses and platform governance. |
AI Governance | Executive Order and sector-specific regulation. | Rights-based oversight and accountability frameworks. | Developing policy framework. |
Predictive Policing | Extensive experimentation and growing scrutiny. | Limited deployment and oversight concerns. | Early-stage adoption. |
Digital Evidence | Federal Rules of Evidence and Case Law. | PACE, Criminal Procedure Rules, and Common Law. | Bharatiya Sakshya Adhiniyam, 2023. |
Data Protection | Sectoral Model. | UK GDPR and Data Protection Act. | Digital Personal Data Protection Act, 2023 |
Source: Compiled by the Authors from Department for Science, Innovation and Technology, A Pro-Innovation Approach to AI Regulation, HM Government, 2023; Regulation (EU) 2024/1689 (Artificial Intelligence Act); Bharatiya Sakshya Adhiniyam, 2023; Digital Personal Data Protection Act, 2023; Federal Rules of Evidence (United States); Information Commissioner's Office & The Alan Turing Institute, Explaining Decisions Made with AI, 2025
The comparative analysis indicates that no jurisdiction has developed a complete response to the challenges posed by artificial intelligence, deepfakes, blockchain technologies, and predictive policing. Across jurisdictions, tensions persist between innovation and accountability, efficiency and fairness, and automation and human oversight. The evidence suggests that the reliability of digital evidence depends less on technological sophistication than on the effectiveness of governance frameworks encompassing authentication procedures, forensic expertise, judicial oversight, transparency, and accountability. The central regulatory challenge, therefore, lies in ensuring that technological innovation remains consistent with fundamental legal principles and the rule of law.
The central finding of this study is that contemporary technologies simultaneously strengthen and destabilize evidentiary systems. Empirical evidence indicates that artificial intelligence can enhance forensic analysis and the processing of large digital datasets, while blockchain-based systems may improve evidentiary preservation, auditability, and chain-of-custody management. Similarly, predictive analytics have introduced new approaches to crime forecasting and resource allocation.[76]
Yet these technologies have also created new vulnerabilities. Generative AI has enabled increasingly sophisticated deepfakes and synthetic media, while predictive policing systems have generated concerns regarding transparency, accountability, and algorithmic bias. Similarly, although blockchain technologies can strengthen evidentiary preservation, they cannot independently verify the authenticity of information entered into a system.[77]
The resulting picture is neither one of straightforward technological progress nor decline. Instead, the evidence demonstrates that the same technologies capable of enhancing verification, evidentiary management, and investigative efficiency can also create new opportunities for manipulation, bias, and uncertainty. Consequently, evidentiary reliability depends not solely upon technological capability but upon the legal institutions, governance structures, organisational practices, and human decision-making processes that shape the development and use of these technologies. This finding is consistent with law-and-technology scholarship, which emphasises that the effects of technological innovation are mediated by the institutional environments within which it operates.[78]
A second finding concerns the widening gap between technologies used to authenticate evidence and those capable of generating fabricated content. Historically, photographs, audio recordings, and videos were generally regarded as reliable forms of proof despite the possibility of manipulation. Advances in artificial intelligence, however, have significantly reduced the barriers to creating convincing synthetic media. Deepfakes, cloned voices, and AI-generated images increasingly challenge traditional assumptions regarding the authenticity of digital evidence.[79]
In response, authentication technologies have continued to evolve. Deepfake-detection systems, forensic analysis tools, cryptographic verification mechanisms, and blockchain-based recordkeeping have been developed to strengthen evidentiary reliability. However, the relationship between generation and detection increasingly resembles an arms race, with advances in synthetic-media creation frequently matched by corresponding improvements in verification technologies.[80]
The implications for criminal justice systems are significant. As digital evidence becomes more susceptible to sophisticated forms of manipulation, authentication has become increasingly complex and dependent upon specialised expertise. Courts therefore rely more heavily on forensic analysis, metadata examination, cryptographic verification, and related techniques when assessing digital evidence. Consequently, the evidentiary value of a digital record depends less on its apparent authenticity than on the reliability of the processes used to verify its origin and integrity.
A third finding is the gradual emergence of hybrid verification models. Evidence from multiple jurisdictions suggests that neither technological solutions nor traditional legal mechanisms alone are sufficient to address contemporary evidentiary challenges. Criminal justice institutions increasingly combine technological and procedural safeguards, using blockchain for chain-of-custody preservation, AI-assisted tools for detecting manipulation, forensic expertise for authenticity assessment, and judicial procedures for evaluating reliability. Within these frameworks, technology supports decision-making, but responsibility for evidentiary and operational judgments remains with human actors.[81] The concept of hybrid verification thus reflects the growing integration of technological systems within established legal and institutional processes. Rather than displacing traditional safeguards, emerging technologies increasingly operate alongside forensic expertise, judicial oversight, and procedural protections to support evidentiary assessment.[82]
This finding challenges deterministic assumptions frequently encountered in discussions of emerging technologies. Technology is neither a substitute for law nor merely an external threat to legal systems. Instead, its effects are shaped by the institutional and legal frameworks within which it is deployed.[83] Accordingly, future reforms should focus not on choosing between human judgment and technological assistance, but on developing governance frameworks that enable both to operate in a complementary and mutually reinforcing manner. Effective regulation therefore requires integrating technological innovation with accountability, transparency, and procedural fairness.
Perhaps the most significant finding concerns institutional capacity. Comparative evidence suggests that the effectiveness of artificial intelligence, blockchain systems, and predictive algorithms depends less on technological sophistication than on the institutions governing them. Jurisdictions with stronger forensic infrastructures, clearer evidentiary standards, specialised expertise, and robust oversight mechanisms are generally better positioned to manage technological risks. Experiences with predictive policing demonstrate that concerns regarding bias often arise from historical datasets and inadequate oversight rather than from algorithms alone. [84] Similarly, blockchain systems cannot ensure authenticity where collection procedures are flawed, and deepfake-detection technologies cannot compensate for deficiencies in forensic capacity.[85]
India's experience is particularly instructive in this regard. The Bharatiya Sakshya Adhiniyam, 2023, represents an important step toward modernizing evidentiary law and adapting legal frameworks to digital forms of evidence. However, the comparative evidence indicates that legal reform is only one component of effective governance. The successful management of synthetic media and algorithmic systems also requires trained investigators, technologically equipped forensic laboratories, judicial familiarity with digital evidence, and continuous institutional adaptation.[86]
The broader implication is that technology should be viewed as an enabling rather than a determining factor in evidentiary reliability. Although technological tools can enhance investigative and evidentiary capabilities, their effectiveness ultimately depends upon the institutional environments within which they operate. Institutions therefore remain the critical variable in determining whether technological innovation strengthens or undermines the administration of justice.
The findings permit a broader reassessment of reliability and admissibility within digital evidentiary systems. Reliability is increasingly shaped by competing technological forces: AI-assisted forensic tools may strengthen evidentiary assessment, while deepfakes undermine confidence in digital records; blockchain enhances evidentiary integrity but not authenticity; and predictive systems remain vulnerable to bias and error. Although existing admissibility frameworks have generally proved adaptable, emerging technologies are placing increasing pressure on courts to develop more sophisticated approaches to authentication, expert testimony, and algorithmic transparency.[87]
These developments also raise broader questions of legitimacy, understood as public confidence in legal institutions and decision-making processes. Legitimacy is perhaps the most difficult dimension to measure and the easiest to overlook. Yet even highly accurate technological systems may encounter resistance if they are perceived as opaque, unfair, or unaccountable. Public trust depends not only upon the accuracy of outcomes but also upon the transparency and fairness of the procedures through which those outcomes are produced.[88]
Accordingly, reliability, admissibility, and legitimacy should not be viewed as separate concerns. Reliability influences admissibility, while admissibility in turn affects public confidence in legal institutions. Technological developments increasingly shape all three dimensions simultaneously, making it essential for criminal justice systems to balance technical innovation with transparency, accountability, and procedural fairness.
The findings of this study do not resolve all issues surrounding digital evidence and emerging technologies but identify several questions that will shape future debates, including the authentication of synthetic media, the evaluation of algorithmically generated evidence, the transparency of technological decision-making, and the governance of blockchain-based systems. The evidence suggests that the future of digital evidence will depend not only on technological innovation but also on the capacity of legal institutions to adapt and preserve public confidence. Finally, the central challenge is institutional: ensuring that technological advancement remains compatible with accountability, fairness, transparency, and the rule of law.
The findings suggest that the principal challenge facing contemporary criminal justice systems is not a lack of technological capability but the uneven development of the institutional capacity required to govern it. While artificial intelligence, blockchain authentication, deepfake-detection tools, and predictive analytics can enhance investigative efficiency and evidentiary management, they also create risks relating to authenticity, transparency, accountability, and procedural fairness. Policy responses should therefore prioritise institutional readiness, ensuring that technological innovation remains consistent with constitutional principles, evidentiary standards, and due-process guarantees.[89]
The emergence of synthetic media necessitates a re-evaluation of traditional evidentiary assumptions. Courts can no longer rely solely upon the apparent authenticity of photographs, audio recordings, and video files. As deepfakes and other forms of AI-generated content become increasingly sophisticated, the risk of fabricated material entering investigative and judicial processes has grown substantially.[90]
Enhanced authentication standards are therefore becoming increasingly important for high-risk categories of digital evidence. Potential safeguards include mandatory metadata preservation, cryptographic hashing at the point of collection, standardized chain-of-custody procedures, independent forensic verification of contested evidence, and judicial guidelines addressing synthetic media. The findings suggest that evidentiary frameworks should move beyond presumptions of authenticity toward demonstrable verification, thereby strengthening confidence in digital evidence.
Blockchain technologies offer valuable opportunities for strengthening evidentiary preservation and chain-of-custody management. However, the findings of this study indicate that blockchain should not be treated as a self-sufficient solution to evidentiary challenges. While the technology can preserve records with considerable integrity, it cannot independently verify the authenticity of information entered into a system.[91]
Policy initiatives should therefore integrate blockchain within broader verification frameworks that combine technological safeguards with human oversight. Recommended measures include blockchain-based evidence management systems for forensic laboratories, secure timestamping mechanisms for electronic records, independent audits of blockchain platforms, and interoperability standards supporting cross-jurisdictional verification. Such measures can strengthen evidentiary integrity and transparency while ensuring that blockchain functions as a complementary component of evidentiary governance rather than a substitute for forensic expertise and procedural safeguards.
Predictive policing programmes and AI-assisted investigative tools illustrate the importance of algorithmic accountability. The experiences of PredPol, the Strategic Subject List, and related initiatives demonstrate that data-driven systems may unintentionally reproduce biases embedded within historical datasets.[92]
To address these concerns, governments should establish mechanisms that promote transparency, auditability, and explainability. Recommended safeguards include mandatory algorithmic impact assessments, independent evaluations of bias and accuracy, public reporting of system performance, periodic reviews of predictive policing programmes, and accessible avenues for challenging algorithmically informed decisions. The evidence further suggests that such technologies should operate as decision-support tools rather than autonomous decision-makers, with meaningful human oversight remaining central to criminal justice governance.
Technological sophistication alone cannot compensate for deficiencies in institutional expertise. As courts increasingly encounter evidence involving artificial intelligence, digital forensics, blockchain verification, and algorithmic analysis, many legal systems continue to face shortages of specialised technical knowledge. Strengthening professional capacity is therefore essential to the effective evaluation of technologically derived evidence.[93]
Governments should enhance institutional capacity through judicial education on digital evidence, specialised forensic training, certification frameworks for digital forensic practitioners, interdisciplinary collaboration between legal and technical experts, and dedicated digital-evidence units within law-enforcement agencies. The objective is not to make all decision-makers technical specialists, but to ensure that they possess the expertise necessary to assess digital evidence critically, responsibly, and in accordance with legal standards.
The comparative analysis undertaken in this study reveals a growing international movement toward the regulation of artificial intelligence. The European Union AI Act, regulatory initiatives in the United Kingdom, and emerging policy developments in the United States reflect increasing recognition that AI systems require specialised governance frameworks.[94] India, on the other hand, has taken important legislative steps through the Bharatiya Sakshya Adhiniyam, 2023, and the Digital Personal Data Protection Act, 2023. However, the expansion of AI-assisted decision-making highlights the need for a more comprehensive approach to accountability, transparency, and evidentiary reliability. Potential measures include risk-based regulation, sector-specific oversight mechanisms, independent supervisory authorities, mandatory disclosure requirements for high-risk AI systems, and greater stakeholder participation. Such initiatives could strengthen public trust while supporting responsible technological innovation.
Digital evidence increasingly transcends national boundaries, particularly in investigations involving cybercrime, online fraud, misinformation, and transnational criminal networks. Consequently, effective governance requires international cooperation alongside domestic reform. Priority areas include cross-border evidence sharing, harmonization of evidentiary standards, joint forensic research, collaborative deepfake-detection initiatives, and information-sharing mechanisms addressing emerging technological threats.[95] Existing frameworks such as the Convention on Cybercrime (Budapest Convention, 2001) provide a useful foundation for these efforts. However, the continuing evolution of artificial intelligence and synthetic media highlights the need for deeper collaboration to ensure that legal systems can effectively address emerging digital challenges.
Finally, the policy challenge is not whether criminal justice systems should adopt technological innovation, but how these technologies can be integrated within institutions that remain transparent, accountable, and consistent with democratic values. The evidence examined in this study suggests that trustworthy digital justice systems require a balance between innovation and oversight. While artificial intelligence, blockchain, and predictive analytics can enhance investigations, evidentiary integrity, and resource allocation, their effectiveness ultimately depends upon robust legal safeguards, professional expertise, and institutional accountability. Future reforms should therefore focus on strengthening the governance frameworks that shape the use of emerging technologies rather than on technological adoption alone.
The analysis undertaken in this study provides clear answers to the principal and subsidiary research questions and supports the central hypothesis advanced at the outset of the study. In response to the principal research question, the study finds that artificial intelligence, deepfakes, blockchain authentication systems, and predictive policing technologies exert a complex and often contradictory influence on the reliability, admissibility, and legitimacy of digital evidence. While AI-assisted forensic analysis, blockchain-based record preservation, and predictive analytics can enhance investigative efficiency and evidentiary management, deepfakes, algorithmic opacity, and embedded biases simultaneously generate new risks for evidentiary authenticity, procedural fairness, and institutional legitimacy.
In relation to the first subsidiary research question, the study finds that blockchain-based authentication systems can significantly strengthen evidentiary integrity through tamper-resistant recordkeeping, chain-of-custody verification, and auditability. However, blockchain cannot independently verify the authenticity or accuracy of information at the point of entry and therefore functions as a mechanism for preserving integrity rather than guaranteeing authenticity.
In relation to the second subsidiary research question, the study finds that existing legal and regulatory frameworks in the United States, the United Kingdom, and India are beginning to adapt to the challenges posed by AI-generated and algorithmically mediated evidence. Nevertheless, these frameworks remain only partially equipped to address rapidly evolving technological risks and therefore require continuing adaptation through transparency, explainability, procedural safeguards, and judicial oversight.
Collectively, these findings support the study’s central hypothesis that technological innovation does not uniformly enhance evidentiary reliability. Rather, reliability depends upon the interaction of verification mechanisms, institutional safeguards, forensic expertise, algorithmic transparency, and legal oversight.
The rapid integration of artificial intelligence, blockchain technologies, and predictive analytics into criminal justice systems has transformed the evidentiary landscape. Digital evidence now plays a central role in investigations and judicial proceedings, creating opportunities for enhanced investigative efficiency, preservation of evidence, and informed decision-making. Examined through the lenses of evidentiary reliability, algorithmic governance, and digital trust, the analysis demonstrates that artificial intelligence functions both as a verification tool and a source of evidentiary uncertainty.[96]
The discussion of blockchain technologies revealed a similarly nuanced picture. Blockchain-based systems offer significant potential for strengthening chain-of-custody management, evidentiary preservation, and auditability. However, evidentiary integrity should not be conflated with authenticity. While blockchain can verify that information has remained unchanged after entry into a system, it cannot independently establish the accuracy of that information at the point of collection. Consequently, blockchain is best understood as a component of broader verification frameworks rather than a comprehensive solution to contemporary evidentiary challenges.[97]
The examination of predictive policing highlighted the governance challenges associated with algorithmic decision-making. Experiences involving PredPol, the Strategic Subject List, and similar initiatives demonstrate that predictive systems often inherit the strengths and limitations of the underlying data. Although predictive analytics may improve resource allocation and support investigative functions, concerns regarding bias, transparency, and accountability persist. The evidence suggests that effective governance depends upon continuous oversight, explainability, independent review, and meaningful human involvement in algorithmically informed decision-making processes.[98]
The comparative analysis of the United States, the United Kingdom, and India demonstrates that technological innovation alone is insufficient to ensure evidentiary reliability. Despite adopting different regulatory approaches, all three jurisdictions reveal a common lesson: institutional capacity is as important as technological capability. Courts, forensic laboratories, investigators, policymakers, and oversight bodies play a decisive role in determining whether emerging technologies enhance or undermine evidentiary reliability, procedural fairness, and the administration of justice.[99]
A central finding of this study is that the future of digital evidence will be shaped less by technology itself than by the quality of institutional governance surrounding its use. Reliability, admissibility, and legitimacy increasingly depend upon hybrid frameworks that combine technological safeguards with legal standards, forensic expertise, judicial scrutiny, and public accountability. As artificial intelligence and digital ecosystems continue to evolve, criminal justice institutions will face growing pressures to adapt. Future research should therefore examine the implications of generative AI, automated evidentiary assessment, explainable AI, and cross-border digital investigations, while encouraging interdisciplinary collaboration among legal, technical, forensic, and policymakers. Finally, the enduring challenge is not technological adoption but ensuring that innovation remains compatible with fairness, transparency, accountability, and the rule of law.
National Institute of Standards and Technology (NIST), Digital Evidence, available at: https://www.nist.gov/digital-evidence (last visited 12 June 2026). ↑
Digital Evidence Emerging as Decisive Tool in Trials: Ex-judge, The Times of India, 20 August 2025, available at: https://timesofindia.indiatimes.com/city/lucknow/digital-evidence-emerging-as-decisive-tool-in-trials-ex-judge/articleshow/123395039.cms (last visited 12 June 2026). ↑
Yisroel Mirsky & Wenke Lee, The Creation and Detection of Deepfakes: A Survey (2020), available at: https://arxiv.org/abs/2004.11138 (last visited 12 June 2026). ↑
Haiying Guan, James Horan & Andrew Zhang, Guardians of Forensic Evidence: Evaluating Analytic Systems Against AI-Generated Deepfakes, NIST (2025), available at: https://www.nist.gov/publications/guardians-forensic-evidence-evaluating-analytic-systems-against-ai-generated-deepfakes (last visited 12 June 2026). ↑
How AI Is Revolutionizing Digital Forensics, Police Chief Magazine, available at:https://www.policechiefmagazine.org/ai-is-revolutionizing-digital-forensics-magnet (last visited 12 June 2026). ↑
Paula Fraga-Lamas & Tiago M. Fernández-Caramés, Fake News, Disinformation, and Deepfakes: Leveraging Distributed Ledger Technologies and Blockchain to Combat Digital Deception and Counterfeit Reality (2024), available at: https://arxiv.org/abs/1904.05386 (last visited 12 June 2026) ↑
D. J. Varghese, Electronic Form Evidence under the Bharatiya Sakshya Adhiniyam, 2023 (2024), SSRN Electronic Journal, available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4869744 (last visited 12 June 2026). ↑
Eoghan Casey, Digital Evidence and Computer Crime: Forensic Science, Computers and the Internet, 3rd edn., Academic Press, Burlington, 2011, available at: https://shop.elsevier.com/books/digital-evidence-and-computer-crime/casey/978-0-08-092148-8 (last visited 12 June 2026). ↑
Christopher Kuner, Transborder Data Flows and Data Privacy Law, Oxford University Press, Oxford, 2013, available at: https://tind.wipo.int/record/48005 (last visited 12 June 2026). ↑
Stuart Russell & Peter Norvig, Artificial Intelligence: A Modern Approach, 4th edn., Pearson, Hoboken, 2021, available at: https://aima.cs.berkeley.edu/ (last visited 12 June 2026). ↑
Europol, Facing Reality? Law Enforcement and the Challenge of Deepfakes, 2022, available at: https://www.europol.europa.eu/publications-events/publications/facing-reality-law-enforcement-and-challenge-of-deepfakes (last visited 12 June 2026). ↑
Cathy O'Neil, Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy, Crown Publishing, New York, 2016; Ruha Benjamin, Race After Technology, Polity Press, Cambridge, 2019; Virginia Eubanks, Automating Inequality, St. Martin's Press, New York, 2018. ↑
Frank Pasquale, The Black Box Society: The Secret Algorithms That Control Money and Information, Harvard University Press, Cambridge, 2015, available at: https://www.hup.harvard.edu/books/9780674970847 (last visited 12 June 2026). ↑
Yisroel Mirsky & Wenke Lee, “The Creation and Detection of Deepfakes: A Survey”, (2021) ACM Computing Surveys, Vol. 54, No. 1, available at: https://dl.acm.org/doi/10.1145/3425780 (last visited 12 June 2026). ↑
World Economic Forum, Global Risks Report 2024, Geneva, 2024, available at: https://www.weforum.org/publications/global-risks-report-2024/ (last visited 12 June 2026). ↑
Robert Chesney & Danielle Citron, “Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security”, (2019) 107 California Law Review 1753, available at: https://www.californialawreview.org/print/deep-fakes-a-looming-challenge-for-privacy-democracy-and-national-security/ (last visited 12 June 2026). ↑
Danielle K. Citron, The Fight for Privacy: Protecting Dignity, Identity and Love in the Digital Age, W.W. Norton, New York, 2022. ↑
Satoshi Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System” (2008), available at: https://bitcoin.org/bitcoin.pdf (last visited 12 June 2026). ↑
Melanie Swan, “Blockchain: Blueprint for a New Economy,” O'Reilly Media, Sebastopol, 2015. ↑
European Commission, European Blockchain Observatory and Forum, BLOCKCHAIN OBSERVATORY, available at: https://blockchain-observatory.ec.europa.eu/index_en ↑
Angela Walch, “In Code(rs) We Trust: Software Developers as Fiduciaries in Public Blockchains”, in Regulating Blockchain,” Oxford University Press, 2019, available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3203198 (last visited 12 June 2026). ↑
Andrew Guthrie Ferguson, The Rise of Big Data Policing: Surveillance, Race, and the Future of Law Enforcement, NYU Press, New York, 2017, available at: https://lawcat.berkeley.edu/record/94720 (last visited 12 June 2026). ↑
RAND Corporation, Predictive Policing: The Role of Crime Forecasting in Law Enforcement Operations, 2013, available at: https://www.rand.org/pubs/research_reports/RR233.html (last visited 12 June 2026). ↑
Sarah Brayne, Predict and Surveil: Data, Discretion, and the Future of Policing, Oxford University Press, Oxford, 2020, available at: https://global.oup.com/academic/product/predict-and-surveil-9780190684099(last visited 12 June 2026). ↑
Ronald J. Allen & Alex Stein, “Evidence, Probability, and the Burden of Proof”, (2013) 55(3) Arizona Law Review 557–602, available at: https://journals.librarypublishing.arizona.edu/arizlrev/article/id/7128/ (last visited 12 June 2026). ↑
Karen Yeung, “Algorithmic Regulation: A Critical Interrogation”, (2018) 12(4) Regulation & Governance 505-523, available at: https://onlinelibrary.wiley.com/doi/10.1111/rego.12158 (last visited 12 June 2026). DOI: 10.1111/rego.12158. ↑
Ruha Benjamin, Race After Technology: Abolitionist Tools for the New Jim Code, Polity Press, Cambridge, 2019, available at: https://www.politybooks.com/bookdetail?book_slug=race-after-technology--9781509526406 (last visited 12 June 2026). ↑
Frank Pasquale, supra note 13 ↑
European Commission, European Blockchain Observatory and Forum, supra note 20. ↑
Stuart Russell & Peter Norvig, supra note 10. ↑
Europol, supra note 11. ↑
National Institute of Standards and Technology (NIST), Digital Evidence and Forensic Science Activities, available at: https://www.nist.gov/topics/forensic-science (last visited 12 June 2026). ↑
Interpol, Artificial Intelligence Toolkit for Responsible AI Innovation in Law Enforcement, 2023, available at: https://www.interpol.int/en/How-we-work/Innovation/Artificial-Intelligence (last visited 12 June 2026). ↑
Europol, Internet Organised Crime Threat Assessment (IOCTA) 2024, available at: https://www.europol.europa.eu/publication-events/main-reports/internet-organised-crime-threat-assessment-iocta-2024 (last visited 12 June 2026). ↑
Clare Garvie, Written Testimony before the House Committee on Oversight and Reform, 22 May 2019, available at: https://www.congress.gov/116/meeting/house/109521/witnesses/HHRG-116-GO00-Wstate-GarvieC-20190522.pdf. ↑
Yisroel Mirsky & Wenke Lee, supra note 14. ↑
Robert Chesney & Danielle Citron, supra note 16. ↑
Company worker in Hong Kong pays out £20m in deepfake video call scam, 2024. https://www.theguardian.com/world/2024/feb/05/hong-kong-company-deepfake-video-conference-call-scam ↑
Robert Chesney & Danielle Citron, supra note 16. ↑
Haiying Guan, James Horan & Andrew Zhang, supra note 4. ↑
European Union Artificial Intelligence Act, Regulation (EU) 2024/1689, available at: https://eur-lex.europa.eu. ↑
D. J. Varghese, supra note 7. ↑
Satoshi Nakamoto, supra note 18. ↑
Melanie Swan, supra note 19. ↑
Primavera De Filippi & Aaron Wright, Blockchain and the Law: The Rule of Code, Harvard University Press, Cambridge, 2019, available at: https://www.hup.harvard.edu/books/9780674241596 ↑
Angela Walch, supra note 21. ↑
Eoghan Casey, supra note 8. ↑
e-Estonia Briefing Centre, Frequently Asked Questions about Estonian Blockchain Technology, available at: https://e-estonia.com/wp-content/uploads/faq_estonian_blockchain_technology.pdf ↑
Supreme People's Court of China, Provisions on Several Issues Concerning the Trial of Cases by Internet Courts, 2018, available at: https://www.chinalawtranslate.com/en/the-supreme-peoples-courts-provisions-on-several-issues-related-to-trial-of-cases-by-the-internet-courts/ ↑
European Commission, European Blockchain Observatory and Forum, supra note 20. ↑
European Commission, European Blockchain Services Infrastructure (EBSI), available at: https://ec.europa.eu/digital-building-blocks (last visited 12 June 2026). ↑
Angela Walch, “The Path of the Blockchain Lexicon (and the Law)”, (2017) 36 Review of Banking and Financial Law 713, available through Boston University School of Law repository. available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2940335 ↑
Andrew Guthrie Ferguson, supra note 22. ↑
Lawrence W. Sherman, “Hot Spots of Crime and Criminal Careers of Places”, in David Weisburd & John Eck (eds.), Crime and Place, Vol. 4, Criminal Justice Press, New York, 1995, available at <https://popcenter.asu.edu/sites/g/files/litvpz3631/files/library/CrimePrevention/Volume_04/02-Sherman-2.pdf> (last visited 12 June 2026) ↑
Walter L. Perry et al., Predictive Policing: The Role of Crime Forecasting in Law Enforcement Operations, RAND Corporation, Santa Monica, 2013, available at: https://www.rand.org/pubs/research_reports/RR233.html ↑
Ibid. ↑
Kristian Lum & William Isaac, “To Predict and Serve?”, (2016) 13 Significance 14-19, available at: https://rss.onlinelibrary.wiley.com/doi/10.1111/j.1740-9713.2016.00960.x ↑
Review of Selected Los Angeles Police Department Data-Driven Policing Strategies, Report No. BPC 19-0072, 12 March 2019, available at: https://www.lapdpolicecom.lacity.org/031219/BPC_19-0072.pdf ↑
Andrew Guthrie Ferguson, supra note 22. ↑
Jessica Saunders, Priscillia Hunt & John S. Hollywood, “Predictions Put into Practice: A Quasi-Experimental Evaluation of Chicago's Predictive Policing Pilot”, (2016) 12(3) Journal of Experimental Criminology 347-371, available at: https://www.rand.org/pubs/external_publications/EP67204.html; Andrea L. DaViera, Marbella Uriostegui, Aaron Gottlieb & Ogechi Cynthia Onyeka, “Risk, Race, and Predictive Policing: A Critical Race Theory Analysis of the Strategic Subject List”, (2024) 73(1-2) American Journal of Community Psychology 91-103, available at: https://onlinelibrary.wiley.com/doi/10.1002/ajcp.12671 ↑
Sarah Brayne, supra note 24. ↑
Cathy O'Neil, Weapons of Math Destruction, Crown Publishing, New York, 2016; Ruha Benjamin, Race After Technology, Polity Press, Cambridge, 2019. ↑
Frank Pasquale, supra note 13. ↑
Federal Rules of Evidence, U.S. Congress, available at: https://www.law.cornell.edu/rules/fre (last visited 12 June 2026). ↑
Danielle K. Citron & Robert Chesney, “Deep Fakes and the New Disinformation War”, (2019) 107 Foreign Affairs 147. https://www.law.virginia.edu/scholarship/publication/danielle-k-citron/1157056 ↑
The White House, Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, 30 October 2023, available at: https://www.federalregister.gov/documents/2023/11/01/2023-24283/safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence (last visited 12 June 2026). ↑
Review of Selected Los Angeles Police Department Data-Driven Policing Strategies, Report No. BPC 19-0072, 12 March 2019, available at: https://www.lapdpolicecom.lacity.org/031219/BPC_19-0072.pdf ↑
Department for Science, Innovation and Technology, A Pro-Innovation Approach to AI Regulation (White Paper), HM Government, 2023, available at: https://www.gov.uk/government/publications/ai-regulation-a-pro-innovation-approach (last visited 12 June 2026). ↑
Government Digital Service, Department for Science, Innovation and Technology, Information Commissioner's Office and The Alan Turing Institute, Explaining Decisions Made with AI, 27 January 2025, available at:https://www.gov.uk/data-ethics-guidance/explaining-decisions-made-with-ai (last visited 12 June 2026). ↑
R (Bridges) v Chief Constable of South Wales Police [2020] EWCA Civ 1058. https://www.matrixlaw.co.uk/2026/02/18/r-bridges-v-south-wales-police-2020-ewca-civ-1058/ ↑
National Crime Records Bureau, Crime in India Reports and related digital policing initiatives. https://www.ncrb.gov.in/ ↑
Bharatiya Sakshya Adhiniyam, 2023. https://www.mha.gov.in/sites/default/files/2024-04/250882_english_01042024_0.pdf ↑
Anvar P.V. v. P.K. Basheer (2014) 10 SCC 473; Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal (2020) 7 SCC 1. ↑
Bureau of Police Research and Development, Government of India, Smart Policing Initiatives in India. https://bprd.nic.in/page/compilation_of_best_practices_in_smart_policing ↑
Digital Personal Data Protection Act, 2023. https://www.meity.gov.in/static/uploads/2024/06/2bf1f0e9f04e6fb4f8fef35e82c42aa5.pdf ↑
Walter L. Perry et al., Predictive Policing: The Role of Crime Forecasting in Law Enforcement Operations, RAND Corporation, 2013. ↑
Yisroel Mirsky & Wenke Lee, supra note 14. ↑
Frank Pasquale, supra note 13. ↑
Yisroel Mirsky & Wenke Lee, supra note 14. ↑
Robert Chesney & Danielle K. Citron, supra note 16. ↑
Andrew Guthrie Ferguson, supra note 22. ↑
Robert Chesney & Danielle K. Citron, supra note 16. ↑
Frank Pasquale, supra note 13. ↑
Andrea L. DaViera, Marbella Uriostegui, Aaron Gottlieb & Ogechi Cynthia Onyeka, “Risk, Race, and Predictive Policing: A Critical Race Theory Analysis of the Strategic Subject List”, (2024) 73(1-2) American Journal of Community Psychology 91-103. ↑
Angela Walch, supra note 21. ↑
D. J. Varghese, supra note 7. ↑
Arjun Panditrao Khotkar v. Kailash Kushanrao Gorantyal, (2020) 7 SCC 1 ↑
Department for Science, Innovation and Technology, supra note 68. ↑
Karen Yeung, supra note 26. ↑
Robert Chesney & Danielle K. Citron, supra note 16. ↑
Angela Walch, supra note 21. ↑
Andrea L. DaViera et al., supra note 84. ↑
D. J. Varghese, supra note 7. ↑
Department for Science, Innovation and Technology, supra note 68. ↑
Council of Europe, Convention on Cybercrime (Budapest Convention), ETS No. 185, available at: https://www.coe.int/en/web/cybercrime/the-budapest-convention (last visited 12 June 2026). ↑
Robert Chesney & Danielle K. Citron, supra note 16. ↑
Angela Walch, supra note 21. ↑
Andrew Guthrie Ferguson, supra note 22. ↑
D. J. Varghese, supra note 7. ↑