Abstract
The proliferation of artificial intelligence (‘AI’) surveillance technologies that utilise machine learning (‘ML’) capabilities to gather and process information to derive insights about individuals has transformed the surveillance landscape by enabling governments and private entities to engage in increasingly pervasive monitoring and data-driven profiling. In India, this transformation is compounded by the presence of executive-controlled surveillance mechanisms under the legal framework of the Telecommunications Act 2023, the Telecommunications (Procedures and Safeguards for Lawful Interception of Messages) Rules 2024, and the Digital Personal Data Protection Act 2023, amongst others, which authorises wide-ranging data interception and dragnet surveillance while exempting government agencies from any meaningful oversight.
This framework, which is anchored in expansive executive powers, opaque data practices and procedures and minimal judicial oversight, and enabled by AI technologies such as facial recognition and predictive policing tools, disproportionately targets marginalised populations due to the inherent risk of algorithmic bias and the non-procedural nature of ML/AI systems, which creates a ‘black box’ effect. The fact that AI-generated outputs derived from such opaque systems are increasingly introduced as evidence in criminal proceedings without clear standards for reliability, admissibility, or constitutional compliance may result in evidentiary abuse. This also has ramifications on due process, procedural fairness and algorithmic accountability within the criminal justice system.
This article critically examines India’s evolving AI-driven surveillance regime and its impact on privacy, due process, and algorithmic accountability within India’s criminal justice system. It argues that the lack of judicial oversight, procedural safeguards, and transparency mechanisms renders India’s surveillance framework incompatible with constitutional protections and global standards. The article advocates for integrated privacy and criminal law reforms that prioritise fairness, accountability and transparency across India’s surveillance regime, and emphasises the need for judicial authorisation, algorithmic accountability, and robust admissibility thresholds for AI-generated outputs in criminal proceedings.
Recommended Citation
Ramaswamy, Samyukta
(2024)
"The Evidence Machine: Rethinking Admissibility and Privacy in India's AI Surveillance State,"
Indian Journal of Law and Technology: Vol. 20:
Iss.
2, Article 4.
DOI: 10.55496/XVOH2915
Available at:
https://repository.nls.ac.in/ijlt/vol20/iss2/4
Digital Object Identifier (DOI)
10.55496/XVOH2915
Included in
Constitutional Law Commons, Courts Commons, Criminal Law Commons, Criminal Procedure Commons, Evidence Commons, Internet Law Commons, Legal Remedies Commons, Legislation Commons, Litigation Commons, Privacy Law Commons