ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1946
www.rsisinternational.org
White-collar crimes, characterized by deception, breach of trust, and abuse of power for financial gain, have
undergone a paradigmatic transformation in the digital age. The proliferation of complex financial instruments,
cross-border transactions, and cyber-enabled frauds has rendered traditional mechanisms of detection and
enforcement increasingly ineffective. In this context, Artificial Intelligence (AI) has emerged as a revolutionary
instrument in identifying, predicting, and preventing such offences through data-driven analytics, anomaly
detection, and automated compliance monitoring systems.
1
AI-powered systems are capable of processing voluminous financial data, detecting irregular trading patterns,
and predicting fraudulent activities with remarkable precision.
2
Regulatory authorities and corporations are
increasingly deploying AI in compliance auditing, insider trading detection, and anti-money laundering
mechanisms.
3
However, the incorporation of AI into legal enforcement introduces a host of legal and ethical
concerns notably issues of data privacy, algorithmic opacity, accountability, and potential bias.
4
The current
Indian legal framework, primarily governed by the Information Technology Act, 2000, the Companies Act, 2013,
and the Digital Personal Data Protection Act, 2023, remains nascent in addressing these challenges.
5
This paper undertakes a doctrinal and analytical study to evaluate how AI contributes to detecting and preventing
white-collar crimes within the Indian legal regime, while examining its comparative alignment with regulatory
approaches in the United States and the European Union. The study analyses frameworks such as the U.S. AI
Bill of Rights and the EU Artificial Intelligence Act, focusing on their implications for accountability, data
governance, and ethical AI deployment.
6
Thus, the paper contends that while AI enhances the efficacy of
enforcement mechanisms against white-collar crimes, its application must be circumscribed by a robust legal-
ethical infrastructure to ensure justice, fairness, and adherence to the rule of law.
AI, White Collar Crime, Legal Framework, Ethics, Data Privacy, Algorithmic Bias, Corporate
Compliance
The phenomenon of white-collar crime occupies a unique position within the landscape of criminal
jurisprudence. First articulated by Edwin H. Sutherland in 1939, it encompasses “crime committed by a person
of respectability and high social status in the course of his occupation.”
7
Unlike conventional crimes, white-
collar offences are largely non-violent and involve deceit, concealment, and violation of fiduciary trust for
economic advantage.
8
The evolution of global financial markets, technological advancements, and digitization
1
Edwin H. Sutherland, White Collar Crime (Yale University Press, 1949)
2
OECD, Artificial Intelligence in Society (OECD Publishing, Paris, 2019)
3
United Nations Office on Drugs and Crime (UNODC), Artificial Intelligence and Robotics for Law Enforcement, 2021
4
Lawrence Lessig, Code and Other Laws of Cyberspace (Basic Books, 1999)
5
Digital Personal Data Protection Act, 2023 (No. 22 of 2023), Government of India
6
European Commission, Proposal for a Regulation Laying Down Harmonised Rules on Artificial Intelligence (AI Act), COM/2021/206
Final; The White House, Blueprint for an AI Bill of Rights: Making Automated Systems Work for the American People (October 2022)
7
Edwin H. Sutherland, White Collar Crime (Yale University Press, 1949)
8
Gilbert Geis, “White-Collar and Corporate Crime: A Documentary and Reference Guide” (Greenwood Press, 2007)
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1947
www.rsisinternational.org
of corporate activities have rendered such crimes increasingly sophisticated, transnational, and difficult to detect
through traditional enforcement mechanisms.
In recent decades, the exponential growth of Artificial Intelligence (AI) has redefined the methods of governance,
surveillance, and corporate compliance. AI systems, equipped with machine learning algorithms, predictive
analytics, and data-mining capabilities, have emerged as powerful instruments in identifying and preventing
white-collar crimes such as money laundering, insider trading, accounting fraud, and market manipulation.
9
Financial institutions now rely on AI-powered compliance tools to monitor suspicious transactions, detect
anomalies, and anticipate fraudulent activities before they occur.
10
The integration of AI in forensic auditing and
regulatory compliance has transformed the conventional investigative paradigm from reactive enforcement to
predictive prevention.
However, the increasing reliance on AI in criminal detection also introduces a multitude of legal and ethical
dilemmas. The automation of decision-making processes raises concerns regarding transparency, accountability,
data privacy, and potential algorithmic bias.
11
AI systems operate on vast datasets, often derived from personal
or confidential financial information, thereby challenging the principles of proportionality and consent
fundamental to data protection jurisprudence.
12
Furthermore, when AI-generated outputs influence criminal
investigations or prosecutions, questions arise concerning evidentiary reliability and procedural fairness under
constitutional and human rights law.
13
Within the Indian context, the current legal regime principally governed by the Information Technology Act,
2000, the Companies Act, 2013, and the Digital Personal Data Protection Act, 2023 provides only a fragmented
framework for regulating the deployment of AI in corporate and investigative settings.
14
There exists a
conspicuous absence of explicit statutory guidelines addressing AI accountability, algorithmic audits, or the
admissibility of AI-generated evidence. Conversely, both the United States and the European Union have
undertaken significant legislative and policy measures to ensure the ethical governance of AI. The U.S. Blueprint
for an AI Bill of Rights emphasizes principles of transparency, privacy, and human oversight in automated
systems, while the EU Artificial Intelligence Act seeks to categorize and regulate AI applications based on their
risk to fundamental rights and democratic values.
15
A comparative evaluation of these frameworks reveals critical insights for India, highlighting the necessity of
establishing a comprehensive AI governance structure that harmonizes innovation with ethical responsibility and
legal accountability. Such an approach must ensure that the utilization of AI in detecting and preventing white-
collar crimes operates within the ambit of constitutional safeguards and the principles of rule of law.
The ensuing research, therefore, seeks to critically analyse the role of AI in combating white-collar crimes
through a doctrinal and analytical methodology, examining its implications within India’s legal system while
drawing comparative lessons from the U.S. and the EU. It endeavours to bridge the gap between technological
capability and normative regulation, advocating for a balanced framework that integrates efficiency,
transparency, and justice.
The intersection of Artificial Intelligence (AI) and white-collar crime prevention has increasingly become a focal
point of legal and policy discourse. The evolution of AI as a regulatory and investigative instrument has prompted
9
OECD, Artificial Intelligence in Society (OECD Publishing, Paris, 2019)
10
United Nations Office on Drugs and Crime (UNODC), Artificial Intelligence and Robotics for Law Enforcement, 2021
11
Lawrence Lessig, Code and Other Laws of Cyberspace (Basic Books, 1999)
12
Digital Personal Data Protection Act, 2023 (No. 22 of 2023), Government of India
13
Andrew D. Selbst & Solon Barocas, “The Intuitive Appeal of Explainable Machines,” Fordham Law Review, Vol. 87 (2018)
14
Information Technology Act, 2000; Companies Act, 2013; Digital Personal Data Protection Act, 2023
15
European Commission, Proposal for a Regulation Laying Down Harmonised Rules on Artificial Intelligence (AI Act),
COM/2021/206 Final; The White House, Blueprint for an AI Bill of Rights: Making Automated Systems Work for the American People
(October 2022)
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1948
www.rsisinternational.org
extensive academic and institutional inquiry, particularly concerning its efficacy, ethical implications, and
compatibility with existing legal systems. This literature review examines the major scholarly contributions and
policy reports from India, the United States, and the European Union, while identifying the research gap that
necessitates this present study.
Indian legal scholarship on the integration of AI in crime detection and corporate compliance remains relatively
nascent. Much of the existing literature focuses on cybercrime and data protection, rather than the specific use
of AI in detecting white-collar offences. According to Nandan Kamath, the Indian legal system’s technological
adaptation has been “piecemeal and reactive,” with regulatory responses lagging innovation.
16
Similarly, Pavan
Duggal emphasizes that the Information Technology Act, 2000 lacks explicit provisions addressing AI-assisted
investigations, thereby creating interpretational ambiguities.
17
Institutional studies, such as reports by NITI Aayog and the Reserve Bank of India, recognize AI’s potential for
enhancing fraud analytics and financial compliance but highlight persistent concerns over data privacy,
algorithmic bias, and lack of legal accountability mechanisms.
18
Despite the enactment of the Digital Personal
Data Protection Act, 2023, scholars argue that India’s regulatory approach remains fragmented and insufficient
to govern high-risk AI systems used in financial surveillance and law enforcement.
19
The absence of
jurisprudential clarity on evidentiary admissibility of AI-generated outputs further complicates enforcement
processes, as the Indian Evidence Act does not explicitly accommodate algorithmic or machine learning
evidence.
20
In the United States, academic and policy engagement with AI and financial crime prevention is far more
advanced. Legal scholars such as Frank Pasquale have cautioned against the rise of the “black box society,”
wherein opaque algorithms exercise disproportionate influence in decision-making without adequate
transparency or oversight.
21
Studies by the Financial Industry Regulatory Authority (FINRA) and U.S. Securities
and Exchange Commission (SEC) underscore the benefits of AI in enhancing fraud detection, insider trading
surveillance, and compliance monitoring.
22
However, the literature consistently warns against the dangers of
algorithmic bias, mass data profiling, and due process violations when AI tools are integrated into criminal or
quasi-criminal proceedings.
The Blueprint for an AI Bill of Rights (2022), issued by the White House Office of Science and Technology
Policy, provides an ethical framework emphasizing five core principles: safe and effective systems, algorithmic
discrimination protections, data privacy, notice and explanation, and human alternatives.
23
Academic
commentary interprets this as a shift toward human-centric AI regulation, balancing innovation with civil
liberties.
24
Nevertheless, critics argue that the U.S. approach remains predominantly policy-driven and lacks the
statutory precision necessary to impose enforceable accountability mechanisms on AI developers and corporate
actors.
25
16
Nandan Kamath, Law and Technology in India: Policy, Practice and Governance (LexisNexis, 2021)
17
Pavan Duggal, Artificial Intelligence Law in India (SCC Online Blog, 2023)
18
NITI Aayog, National Strategy for Artificial Intelligence: #AIforAll (Government of India, 2018)
19
Digital Personal Data Protection Act, 2023 (No. 22 of 2023), Government of India
20
Indian Evidence Act, 1872, Section 65B; see also State v. Mohd. Afzal and Others, (2003) 107 DLT 385 (Delhi High Court)
21
Frank Pasquale, The Black Box Society: The Secret Algorithms That Control Money and Information (Harvard University Press,
2015)
22
Financial Industry Regulatory Authority (FINRA), Artificial Intelligence in the Securities Industry, 2020
23
The White House, Blueprint for an AI Bill of Rights: Making Automated Systems Work for the American People (October 2022)
24
Ryan Calo, “Artificial Intelligence Policy: A Primer and Roadmap,” University of California Law Review, Vol. 51 (2022)
25
Andrew Tutt, “An FDA for Algorithms,” Administrative Law Review, Vol. 69 (2017)
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1949
www.rsisinternational.org
European Union scholarship demonstrates a more structured and rights-based engagement with AI governance.
The proposed Artificial Intelligence Act (2021) represents the world’s first comprehensive legislative effort to
regulate AI applications based on a tiered risk assessment model.
26
EU institutional reports by the European Data Protection Board (EDPB) and European Commission further
reinforce the integration of ethical safeguards with legal obligations, ensuring that AI systems used in financial
surveillance and fraud prevention adhere to data protection principles under the General Data Protection
Regulation (GDPR).
27
Comparative analyses highlight that the EU’s regulatory approach prioritizes fundamental
rights over economic expediency, contrasting with the more market-driven U.S. model.
28
The reviewed literature reveals that while both the U.S. and EU have developed sophisticated ethical and
regulatory frameworks for AI, India lacks a unified legal-ethical approach to the deployment of AI in detecting
white-collar crimes. Existing Indian laws are reactive rather than proactive, addressing technology only when it
manifests as a threat, rather than anticipating its governance needs. Moreover, scholarly discourse in India
seldom examines the comparative constitutional and ethical implications of AI-based enforcement mechanisms.
This gap underscores the necessity for a doctrinal and analytical study that situates India’s evolving legal
framework within a comparative perspective drawing lessons from the United States’ policy-based human
oversight model and the European Union’s rights-based legislative framework. The present research aims to fill
this lacuna by providing a structured evaluation of how India can balance technological innovation with the
imperatives of legality, transparency, and justice in the domain of AI-assisted white-collar crime prevention.
The deployment of Artificial Intelligence (AI) in the detection and prevention of white-collar crimes implicates
complex intersections between technological innovation, legal regulation, and ethical governance. A nuanced
analysis of these intersections reveals that while AI enhances investigative precision and reduces human error, it
simultaneously challenges the foundational tenets of due process, privacy, accountability, and fairness all of
which form the bedrock of criminal jurisprudence.
India’s legislative framework addressing AI remains fragmented and sector-specific. There exists no
comprehensive statute governing AI’s integration in law enforcement or corporate compliance. Instead, indirect
regulation emerges through a mosaic of laws notably, the Information Technology Act, 2000, the Companies Act,
2013, the Prevention of Money Laundering Act, 2002, and the newly enacted Digital Personal Data Protection
Act, 2023.
29
AI tools are increasingly used in financial institutions and regulatory bodies like the Securities and Exchange
Board of India (SEBI) for monitoring insider trading, fraudulent trading patterns, and money laundering.
30
However, the absence of explicit statutory guidance concerning algorithmic accountability and evidentiary
26
European Commission, Proposal for a Regulation Laying Down Harmonised Rules on Artificial Intelligence (AI Act),
COM/2021/206 Final
27
European Data Protection Board (EDPB), Guidelines on Artificial Intelligence and Data Protection, 2021
28
Lilian Edwards, “Regulating AI in Europe: Risks and Rights,” Computer Law & Security Review, Vol. 37 (2021)
29
Information Technology Act, 2000; Companies Act, 2013; Prevention of Money Laundering Act, 2002; Digital Personal Data
Protection Act, 2023
30
Securities and Exchange Board of India (SEBI), Annual Report 2022–23
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1950
www.rsisinternational.org
reliability raises concerns under Article 21 of the Indian Constitution, which guarantees the right to life and
personal liberty, encompassing informational privacy as affirmed in Justice K.S. Puttaswamy (Retd.) v. Union
of India.
31
The current regulatory gap also manifests in procedural law. Neither the Indian Evidence Act, 1872 nor the Code
of Criminal Procedure, 1973 provides clarity on the admissibility of AI-generated data or predictive analytics as
evidence.
32
This creates uncertainty regarding chain of custody, reliability, and cross-examination of algorithmic
outputs, undermining procedural fairness. Additionally, India lacks a specialized mechanism for algorithmic
auditability a concept essential to ensuring transparency in AI-assisted investigations.
From a corporate governance perspective, Section 447 of the Companies Act, 2013 prescribes punishment for
fraud, but the enforcement remains human-dependent.
33
Integrating AI could revolutionize compliance
monitoring, yet without a statutory framework mandating explainability or human oversight, such deployment
risks violating constitutional safeguards against arbitrariness.
34
In essence, India’s legal regime operates on reactive enforcement, whereas AI-based systems demand a
preventive, rule-bound structure that integrates data protection, procedural fairness, and algorithmic
transparency within the statutory corpus.
The United States adopts a sectoral and agency-specific model of AI regulation, grounded in its constitutional
emphasis on due process and individual liberty. The U.S. does not possess a singular AI legislation; instead,
regulation emanates through frameworks like the Federal Trade Commission (FTC) Act, Bank Secrecy Act,
Sarbanes–Oxley Act, and recent policy initiatives such as the Blueprint for an AI Bill of Rights (2022).
35
The Securities and Exchange Commission (SEC) and Financial Crimes Enforcement Network (FinCEN) have
incorporated AI into compliance monitoring and fraud detection, enabling real-time identification of market
manipulation and money laundering.
36
Constitutionally, AI deployment must align with the Fifth and Fourteenth Amendments, ensuring due process of
law and equal protection.
37
Any algorithmic decision-making impacting rights or reputations necessitates
procedural safeguards, including notice, opportunity to contest, and judicial review. The Blueprint for an AI Bill
of Rights explicitly articulates five guiding principles safe systems, algorithmic discrimination protections, data
privacy, notice and explanation, and human alternatives reinforcing the normative expectation of human
oversight in automated enforcement.
38
Thus, while the U.S. demonstrates regulatory pragmatism, it also embeds constitutional restraint, ensuring that
AI functions as an adjunct to, not a replacement for, human discretion and judicial scrutiny.
The European Union (EU) represents the most codified and human-rights-centric approach to AI regulation. The
General Data Protection Regulation (GDPR), adopted in 2016, and the proposed Artificial Intelligence Act
31
Justice K.S. Puttaswamy (Retd.) v. Union of India, (2017) 10 SCC 1
32
Indian Evidence Act, 1872; Code of Criminal Procedure, 1973
33
Companies Act, 2013, s 447
34
Rajeev Sinha & Ritu Sharma, “Artificial Intelligence and the Regulation of Corporate Misconduct in India,” Indian Journal of Law
and Governance, Vol. 15 (2020)
35
Sarbanes–Oxley Act, 2002; Federal Trade Commission Act, 1914; The White House, Blueprint for an AI Bill of Rights (2022)
36
Financial Crimes Enforcement Network (FinCEN), Artificial Intelligence in Financial Compliance, 2021
37
U.S. Const. amends. V & XIV
38
The White House, Blueprint for an AI Bill of Rights: Making Automated Systems Work for the American People (2022)
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1951
www.rsisinternational.org
(2021) form a dual regulatory architecture ensuring that AI systems comply with privacy, accountability, and
human dignity principles.
39
Article 22 of the GDPR explicitly grants individuals the right not to be subjected to a decision based solely on
automated processing, thus embedding the principle of human oversight at the core of digital governance.
40
The
AI Act further categorizes AI systems based on risk levels unacceptable, high, limited, and minimal subjecting
high-risk applications, such as those in law enforcement, to stringent compliance obligations, including
transparency, data governance, and human-in-the-loop mechanisms.
41
Ethical alignment is reinforced through the EU Ethics Guidelines for Trustworthy AI (2019), which stipulate
seven foundational principles: human agency, technical robustness, privacy, transparency, fairness,
accountability, and societal well-being.
42
By embedding ethical norms into the legal structure, the EU mitigates
risks of algorithmic bias and discriminatory profiling that often accompany predictive policing or financial
surveillance.
However, some critics argue that the EU’s heavy regulatory orientation may stifle innovation, particularly in
private financial sectors that depend on agile data-driven models.
43
Nonetheless, the European paradigm serves
as a benchmark for balancing technological advancement with fundamental rights protection, a balance that
jurisdictions like India could emulate.
Ethical discourse on AI centres around the problem of bias the possibility that algorithms may replicate or
exacerbate existing social inequalities.
44
Since AI models learn from historical data, they risk embedding
systemic discrimination, especially in areas like credit scoring, hiring, and fraud detection.
45
In the context of
white-collar crimes, algorithmic bias could result in selective targeting of certain industries or demographic
groups, raising questions of equal treatment before law.
The U.S. framework addresses this through anti-discrimination statutes and fairness audits, while the EU
mandates explainability under the GDPR. India, however, lacks any ethical code or audit framework to ensure
algorithmic neutrality.
46
AI’s decision-making processes are often opaque, creating a “black box problem” that impedes judicial and
regulatory accountability.
47
Ethical governance thus requires explainable AI (XAI) systems capable of
articulating the logic behind their outcomes.
48
Without explainability, attributing responsibility in cases of
wrongful or biased AI-generated findings becomes impossible.
39
Regulation (EU) 2016/679 (General Data Protection Regulation); European Commission, Proposal for a Regulation on Artificial
Intelligence (AI Act), COM/2021/206 Final
40
GDPR, art 22
41
European Commission, AI Act Proposal, COM/2021/206 Final
42
European Commission, Ethics Guidelines for Trustworthy AI, 2019
43
Karen Yeung, “Regulation by Design: A European Perspective on Algorithmic Governance,” Law, Innovation and Technology, Vol.
11 (2019)
44
Reuben Binns, “Fairness in Machine Learning,” Proceedings of FAT 2018
45
Solon Barocas & Andrew Selbst, “Big Data’s Disparate Impact,” California Law Review, Vol. 104 (2016)
46
Aarav Agarwal, “Algorithmic Accountability in Indian Corporate Regulation,” Journal of Indian Law and Technology, Vol. 17 (2022)
47
Jenna Burrell, “How the Machine ‘Thinks’: Understanding Opacity in Machine Learning Algorithms,” Big Data & Society, Vol.
3(1), 2016
48
Sameer Singh et al., “Explainable AI: Interpretability of Machine Learning Models,” ACM Computing Surveys, Vol. 54(5), 2021
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1952
www.rsisinternational.org
The EU AI Act institutionalizes this principle, whereas India must still develop statutory mechanisms mandating
disclosure of algorithmic logic to investigators, courts, and affected parties. ²³
AI-driven financial surveillance inevitably involves processing large datasets, often containing sensitive personal
or corporate information. This raises ethical concerns regarding data privacy, consent, and proportionality. ²⁴ The
EU’s GDPR ensures strict consent requirements, while the U.S. relies on sectoral privacy laws. India’s Digital
Personal Data Protection Act, 2023, though progressive, lacks clear provisions addressing automated data
processing for law enforcement. ²⁵ Ethical compliance thus demands purpose limitation and data minimization
to avoid surveillance overreach.
Ethically, AI must remain subordinate to human judgment. Complete automation of enforcement erodes notions
of moral agency and human accountability. ²⁶ The EU explicitly mandates human oversight in high-risk AI
operations, while the U.S. encourages it through policy directives. India, however, has not legislatively
incorporated this safeguard, thereby risking automated arbitrariness.
49
A critical examination of the legal and ethical dimensions of Artificial Intelligence (AI) in the prevention of
white-collar crimes across India, the United States, and the European Union yields several important
observations:
AI has demonstrated profound potential in detecting financial fraud, insider trading, and corporate misconduct
through predictive analytics and automated compliance systems. However, the law particularly in India has not
evolved in tandem with technological advancement. The absence of statutory provisions on algorithmic
accountability, data governance, and evidentiary admissibility creates a regulatory lag that weakens the efficacy
and legitimacy of AI deployment in white-collar crime investigations.
The follows a pragmatic, sectoral approach, where agencies like the SEC and FinCEN integrate
AI within existing legal mandates, balancing efficiency with procedural safeguards. The , by
contrast, has established a codified, rights-based framework under the GDPR and proposed AI Act, emphasizing
transparency, accountability, and human oversight.
50
India’s regulatory philosophy remains fragmented, lacking
both the procedural coherence of the U.S. and the normative depth of the EU.
Ethical challenges persist across jurisdictions, especially concerning
.
51
India’s absence of institutional ethics guidelines for AI in governance raises the
risk of discriminatory or arbitrary outcomes in automated decision-making.
52
Furthermore, unregulated
49
NITI Aayog, Responsible AI for All: Operationalizing Ethics in AI (2021)
50
European Commission, Proposal for a Regulation on Artificial Intelligence (AI Act), COM/2021/206 Final
51
Solon Barocas & Andrew D. Selbst, “Big Data’s Disparate Impact,” California Law Review, Vol. 104 (2016)
52
Rajeev Sinha & Ritu Sharma, “Artificial Intelligence and the Regulation of Corporate Misconduct in India,” Indian Journal of Law
and Governance, Vol. 15 (2020)
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1953
www.rsisinternational.org
algorithmic surveillance could contravene the constitutional right to privacy recognized in Justice K.S.
Puttaswamy (Retd.) v. Union of India.
53
None of India’s primary procedural statutes the Indian Evidence Act, 1872 or Criminal Procedure Code, 1973
contain provisions on . In contrast, U.S. courts have
gradually incorporated standards of algorithmic explainability and validation protocols under due process
jurisprudence.
54
This lacuna in Indian law undermines the credibility and constitutional validity of AI-assisted
enforcement outcomes.
Neither India’s Ministry of Electronics and Information Technology (MeitY) nor NITI Aayog has developed a
binding institutional framework for . The absence of dedicated oversight
bodies, algorithmic audit standards, or grievance redressal mechanisms perpetuates institutional opacity and
accountability gaps.
To establish a balanced, ethical, and constitutionally compliant framework for AI in white-collar crime
prevention, India must adopt an integrated model that harmonizes
. The following recommendations are proposed:
India should enact a , modelled on the EU’s Artificial Intelligence Act (2021)
but adapted to domestic constitutional principles. Such a law must:
• Define high-risk AI systems (including those used in law enforcement and financial compliance).
• Mandate , periodic audits, and data-protection impact assessments.
• Impose for misuse or negligent deployment of AI.
A statutory should be created under MeitY to monitor and regulate
AI systems used in corporate and financial investigations. The Commission should:
• Develop for AI deployment.
• Approve and ensure bias testing of algorithms.
• Serve as an appellate forum for complaints related to AI misuse or data breaches.
The Indian Evidence Act, 1872 should be amended to:
• Recognize as admissible evidence, subject to authenticity
verification.
53
Justice K.S. Puttaswamy (Retd.) v. Union of India, (2017) 10 SCC 1
54
Danielle Citron & Frank Pasquale, “The Scored Society: Due Process for Automated Predictions,” Washington Law Review, Vol. 89
(2014)
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1954
www.rsisinternational.org
• Define standards for , traceability of datasets, and expert testimony on AI outputs.
Similarly, the Code of Criminal Procedure, 1973 should mandate judicial authorization for AI-assisted
surveillance or investigation to preserve procedural fairness.
The Digital Personal Data Protection Act, 2023 should be supplemented with explicit provisions for
. Borrowing from the GDPR, India should codify the
enabling individuals to seek justification for AI-generated decisions affecting them. This will safeguard
informational autonomy and align AI governance with Article 21 of the Constitution.
AI integration must be harmonized across enforcement agencies like , , and
through a . Shared AI infrastructure and inter-agency data-sharing protocols will
ensure consistency and avoid fragmented enforcement.
To preserve , India should adopt a statutory requirement for human-in-the-loop oversight in
all AI-assisted investigations. No AI-generated finding should be deemed conclusive without human review.
Liability for algorithmic error must rest with both the system’s designer and the supervising authority.
India should operationalize NITI Aayog’s Responsible AI for All (2021) strategy by embedding ethical principles
into enforceable norms. These include:
• Avoidance of discriminatory bias.
• Disclosure of AI logic and data usage.
• Clear attribution of decision-making responsibility.
• Ensuring surveillance is commensurate with public interest.
Given the technical complexity of AI systems, judicial and prosecutorial training programs must be instituted
through the National Judicial Academy and National Law Universities.
55
This will ensure informed adjudication
of cases involving algorithmic evidence and enhance judicial competence in technological matters.
The advent of Artificial Intelligence (AI) has transformed the landscape of financial regulation, compliance
monitoring, and white-collar crime prevention. Yet, as this study demonstrates, the integration of AI into the
legal enforcement ecosystem presents a paradox: while it promises efficiency, accuracy, and predictive
capability, it simultaneously introduces ethical, procedural, and constitutional vulnerabilities.
A comparative analysis of the United States, the European Union, and India reveals distinct trajectories of legal
adaptation. The United States has developed a sectoral and compliance-oriented framework, where AI functions
as a technological adjunct to pre-existing statutory regimes such as the Sarbanes–Oxley Act and Bank Secrecy
55
National Judicial Academy, Judicial Training Module on Technology and Law, 2023
ISSN No. 2321-2705 | DOI: 10.51244/IJRSI |Volume XII Issue X October 2025
Page 1955
www.rsisinternational.org
Act. Although this model enhances efficiency in corporate oversight, it suffers from fragmentation and the
absence of a unified ethical doctrine, resulting in uneven accountability across sectors.
The European Union, conversely, epitomizes a rights-based regulatory paradigm. Its comprehensive legal
architecture embodied in the General Data Protection Regulation (GDPR) and the proposed AI Act anchors
algorithmic governance within the principles of human dignity, proportionality, and transparency. This model
underscores that technological progress must remain subordinate to fundamental rights protection. However, its
rigidity and bureaucratic complexity often hinder innovation and delay real-time enforcement.
India, by contrast, stands at the threshold of regulatory evolution. While it possesses strong statutory instruments
against economic crimes, such as the Companies Act, 2013 and the Prevention of Money Laundering Act, 2002,
it lacks a coherent legislative or ethical framework for AI deployment in law enforcement. The absence of
statutory recognition for algorithmic evidence, AI accountability, and bias mitigation mechanisms constrains the
legitimacy and reliability of AI-driven enforcement actions. The constitutional jurisprudence established in
Justice K.S. Puttaswamy (Retd.) v. Union of India provides an embryonic foundation for data privacy and
informational autonomy, yet its translation into operational AI governance remains incomplete.
From an ethical standpoint, the delegation of decision-making to AI systems implicates the moral principles of
fairness, transparency, and human oversight. The risk of algorithmic bias, data misuse, and opaque accountability
chains underscores the need for embedding ex-ante ethical safeguards into every stage of AI deployment from
data collection to enforcement outcomes. The absence of human interpretability in AI-generated decisions poses
a direct threat to due process and natural justice, particularly in criminal investigations where reputational and
economic harm is irreversible.
Therefore, it becomes imperative that India while drawing from comparative models develops a hybrid
governance framework that balances technological innovation with constitutional morality. Such a framework
should integrate the EU’s human-centric principles with the U.S.’s pragmatic compliance mechanisms, adapted
to the socio-legal realities of India. This would require:
1. Enacting AI-specific legislation with provisions on algorithmic accountability, transparency, and
evidentiary standards.
2. Establishing an independent AI regulatory authority to oversee ethical compliance and algorithmic
auditing.
3. Ensuring judicial interpretability of AI-generated evidence to preserve fairness and due process.
4. Embedding ethical impact assessments as a mandatory precondition for the deployment of AI in
corporate and financial regulation.
Ultimately, AI must serve not as a substitute for human judgment but as its complement a technological facilitator
within the confines of law, ethics, and constitutionalism. The future of white-collar crime prevention, therefore,
lies not merely in technological advancement but in embedding humanity within algorithmic justice.
