Source: The Hindu 

Introduction 

The rapid deployment of AI-powered surveillance technologies in India raises critical legal and constitutional red flags. As government agencies increasingly adopt facial recognition and data collection systems, fundamental rights to privacy are under unprecedented threat. The current legal framework fails to provide adequate safeguards, leaving citizens vulnerable to indiscriminate data collection and potential misuse of sensitive personal information. 

What are the Legal Vulnerabilities in AI Surveillance, and How do They Address Security and Privacy Concerns? 

• Substantive Legal Deficiency:  
  • The existing legislative framework demonstrates a profound absence of comprehensive legal provisions that adequately circumscribe the potential constitutional infringements arising from AI-powered surveillance technologies.  
  • The Digital Personal Data Protection Act of 2023 provides broad governmental exemptions that fundamentally undermine individual privacy protections. 
• Constitutional Incongruence:  
  • The deployment of AI surveillance mechanisms directly contravenes the principles established in K.S. Puttaswamy v. Union of India (2017), which explicitly recognized privacy as a fundamental right.  
  • The current regulatory approach fails to implement the proportionality principle, thereby creating a legal vacuum that permits potentially unrestricted data collection and processing. 
• Procedural Inadequacy:  
  • The absence of mandated judicial oversight, transparent consent mechanisms, and specific restrictions on high-risk AI activities creates a systemic legal vulnerability. 
  • This procedural deficit enables potential systematic violations of informational privacy without meaningful legal recourse. 
• Regulatory Arbitrariness:  
  • The lack of a risk-based regulatory framework, in contrast to international standards like the European Union's Artificial Intelligence Act, exposes citizens to unchecked surveillance mechanisms without established legal safeguards or accountability measures. 

How do Different Global Regions Approach AI Surveillance Regulation, and What are Their Legal Strategies and Challenges? 

• European Union (Most Restrictive): 
  • EU AI Act: first regulation on artificial intelligence 
  • Unequivocally prohibits AI technologies deemed as "unacceptable risk" 
  • Legally bans real-time biometric identification in public spaces 
  • Implements stringent risk-categorization framework for AI applications 
  • Imposes substantial financial penalties for non-compliance (up to 35 million euros) 
• United States (Regulatory Emerging): 
  • No comprehensive national AI ban 
  • Implements activity-specific AI governance 
  • President Biden's executive order mandates AI system testing 
  • State-level regulations targeting specific AI applications 
  • Focus on preventing discriminatory AI practices 
• China (State-Controlled Approach): 
  • set of measures for scientific and technical ethics reviews came into effect on 1st December, 2023.  
  • The guidelines for recommendation algorithms began on 1st March, 2024, and are still evolving. 
  • Regulates public-facing AI services extensively 
  • Implements strict governmental oversight on AI development 
  • Controls algorithmic and generative AI services 
  • National strategy emphasizes controlled technological innovation 
• India (Minimal Legal Framework): 
  • Digital Personal Data Protection Act (DPDPA), passed in 2023, was meant to provide a framework for managing consent and ensuring accountability for data privacy in India.  
  • Currently lacks comprehensive AI regulatory mechanism 
  • No explicit legal bans on AI technologies 
  • Deploying AI surveillance without robust legal safeguards 
  • Promised Digital India Act remains unimplemented 
• United Kingdom (Adaptive Regulation): 
  • Adopts "pro-innovation" regulatory approach 
  • Decentralizes AI governance across sector-specific regulators 
  • Emphasizes balanced technological development 
  • Hosts international AI safety discussions 
• South Africa  
  • AI in South Africa is currently governed by laws like the Protection of Personal Information Act (POPIA). 
  • The country has launched the South African Centre for Artificial Intelligence Research (CAIR). 
• Australian 
  • Even without specific legal statutes, the country has introduced voluntary AI ethics principles and guidelines 

What are the Challenges in Using AI in the Judiciary System? 

• Ethical and Bias Concerns 
  • AI systems can inadvertently perpetuate existing biases present in historical legal data 
  • Machine learning models may reflect systemic prejudices found in past judicial decisions 
  • Risk of discriminatory outcomes based on race, gender, socioeconomic background, or other protected characteristics 
• Lack of Transparency and Explainability 
  • Many AI algorithms operate as "black boxes," making it difficult to understand how decisions are reached 
  • Legal systems require clear reasoning and transparent decision-making processes 
  • Difficulty in explaining complex AI-driven legal reasoning to judges, lawyers, and defendants 
• Legal and Accountability Challenges 
  • Unclear legal framework for AI-assisted or AI-driven judicial decisions 
  • Questions about liability when AI systems make errors 
  • Determining responsibility for incorrect or biased legal recommendations 
• Data Quality and Reliability 
  • Legal AI systems depend on high-quality, comprehensive, and unbiased historical legal data 
  • Incomplete or historically skewed legal databases can lead to unreliable predictions 
  • Challenges in maintaining and updating massive legal databases 
• Complexity of Legal Reasoning 
  • Law involves nuanced interpretation, contextual understanding, and ethical considerations 
• AI may struggle with:  
  • Interpreting ambiguous legal language 
  • Understanding subtle contextual details 
  • Applying moral and ethical judgments 
  • Handling unprecedented or unique legal scenarios 
• Privacy and Data Protection 
  • Legal proceedings often involve sensitive personal information 
  • Risks of data breaches and unauthorized access to confidential legal data 
  • Compliance with strict data protection regulations 
• Human Rights and Due Process 
  • Concerns about replacing human judgment with algorithmic decision-making 
  • Potential violation of fundamental rights to a fair trial 
  • Risk of reducing complex human experiences to statistical calculations 
• Technical Limitations 
  • Current AI technologies cannot fully replicate human intuition and empathy 
  • Difficulties in handling complex legal arguments and emotional nuances 
  • Limited ability to understand non-verbal communication and social context 
• Cost and Implementation Challenges 
  • Significant financial investment required for developing sophisticated legal AI systems 
  • Training legal professionals to work effectively with AI technologies 
  • Ongoing maintenance and updating of AI systems 
• Resistance from Legal Professionals 
  • Potential job displacement fears 
  • Professional skepticism about AI's capabilities 
  • Cultural and institutional resistance to technological change 
• Cybersecurity Vulnerabilities 
  • AI systems can be potential targets for hacking or manipulation 
  • Risk of compromising judicial integrity through technological vulnerabilities 
  • Need for robust security measures to protect AI-driven legal systems 

What are Global Bodies and AI Governance Initiatives? 

• Collaborative International Efforts: The United Nations, OECD, and G20 are actively working to establish common principles for AI governance, focusing on critical areas such as: 
  • Ethics 
  • Transparency 
  • Accountability 
  • Fairness 
• Professional Organization Contributions: The World Economic Forum and IEEE are playing significant roles in AI governance, with the IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems developing standards that carry substantial weight in the scientific and technical communities. 
• Multilateral Diplomatic Milestone: The Hiroshima AI Process Friends Group, initiated by Japanese Prime Minister Kishida Fumio, represents a landmark achievement in global AI governance: 
  • Backed by 49 countries (primarily OECD members) 
  • First agreement among G7 democratic leaders 
  • Emphasizes enhancing AI with a focus on protecting individual rights 
• Evolving Regulatory Landscape: The document highlights that AI regulation efforts are: 
  • Ongoing at both global and domestic levels 
  • Characterized by interconnected decision-making 
  • Influenced by ongoing debates and emerging principles 
• Future Regulatory Trends: Countries worldwide are actively discussing and preparing AI regulations, with many expected to develop specific laws in the near future. Examples include: 
  • Egypt unveiling a national AI strategy 
  • Turkey providing data protection guidelines 
  • South Korea rolling out an AI act 
  • Japan establishing AI regulation principles

Conclusion

The unchecked expansion of AI surveillance represents a significant erosion of individual civil liberties in India. Without robust legislative oversight, comprehensive data protection regulations, and clear restrictions on high-risk AI activities, the government risks transforming technological advancement into a tool of systematic privacy invasion. The urgent need is a balanced regulatory approach that prioritizes citizens' constitutional rights while allowing responsible technological innovation.