AI-Powered Surveillance in India

Context: India is at the forefront of adopting AI in surveillance. Technological integration is a welcome move to modernise law enforcement, however, in the absence of suitable legal frameworks it might intersect with constitutional rights of citizens, particularly the right to privacy.

Relevance of the Topic: Mains: AI in surveillance- Challenges, Opportunities, Way Forward

Use of AI in Governance

• In 2019, India announced its ambition to build the world’s largest facial recognition system for policing.
  • Artificial Intelligence (AI)-powered surveillance systems have been deployed across railway stations and the Delhi Police is preparing to use AI for crime patrols. 
• India plans to launch 50 AI-powered satellites to further intensify India’s surveillance infrastructure.

Challenges:

While technological advancements in law enforcement offer potential, they raise significant legal, constitutional, and ethical concerns.

• Privacy Concerns:
  • Right to Privacy: Recognised as a fundamental right under Article 21 in K.S. Puttaswamy vs Union of India (2017).
  • AI-enabled “dragnet surveillance” often involves indiscriminate data collection, infringing on citizens’ informational privacy.
  • Lack of proportional safeguards against AI misuse raises risks of mass surveillance and data breaches.
• Gaps in Legal Frameworks:
  • Digital Personal Data Protection Act (DPDPA), 2023:
    • Grants the government unchecked power to process personal data without the need for consent, when processing data for medical treatment during an epidemic, data related to employment etc.
    • Mandates citizens not to suppress any material information when submitting personal data. This provision (while intended to ensure data accuracy) could lead to punitive measures for something as simple as an outdated address or technical error in data collection systems.
    • Criticised for skewing power towards state surveillance over individual rights.
  • Absence of specific legislation for AI regulation, despite growing deployment of AI-powered systems.
• Unregulated AI Usage:
  • India’s AI surveillance lacks clear guidelines on data collection, processing, storage, and usage and mechanisms to prevent abuse or discrimination.
  • Example: Deployment of facial recognition technologies in Delhi and Hyderabad without public risk assessments or legislative debate.
• Risk of Overreach:
  • International experiences, such as the U.S. Foreign Intelligence Surveillance Act (FISA), shows that surveillance laws can lead to overreach.
  • Expanding India’s AI surveillance infrastructure without sufficient safeguards risks violating constitutional principles of proportionality and legality.

Indian Context:

• India’s surveillance capabilities are growing rapidly with plans for 50 AI-powered satellites and integration of AI in public systems.
• Cases like Telangana Police data breach highlight the misuse of personal data collected through welfare schemes (E.g., “Samagra Vedika”).
• Current frameworks fail to ensure transparency, judicial oversight, or accountability in data collection and AI deployment.

International Context:

• European Union: EU’s Artificial Intelligence Act follows a risk-based approach:
  • Categorises AI applications as unacceptable, high, transparency, or minimal risk.
  • Prohibits real-time biometric identification for law enforcement, except under strict conditions.
• United States: Surveillance laws like FISA offer lessons on the potential for overreach and the need for stringent safeguards.

Impact on Civil Liberties:

AI surveillance, without sufficient safeguards, might risk:

• Privacy violations: Indiscriminate data collection threatens informational privacy.
• Discrimination: Biased AI systems can exacerbate social inequalities.
• Data breaches: Weak safeguards increase vulnerability to cyberattacks.
• Loss of trust: Citizens may lose confidence in public institutions.

Way Forward:

• Regulatory Frameworks:
  • Enact comprehensive legislation for AI governance.
  • Categorize AI applications based on risk levels, with specific restrictions on high-risk activities.
• Transparent Data Practices:
  • Mandate public disclosure of: What data is collected, for what purpose, how it is stored, timelines for data retention and deletion.
  • Narrow exemptions for consent-based data collection, with judicial oversight.
• Independent Oversight:
  • Create independent bodies to oversee AI deployment in public systems.
  • Establish mechanisms for judicial review of surveillance activities.
• International Best Practices:
  • Adopt risk-based regulatory approaches like that of the EU.
• Citizen Awareness:
  • Educate citizens on their privacy rights under the Constitution and relevant laws.
  • Build grievance redressal mechanisms for AI-related violations.
• Strengthen the DPDPA:
  • Narrow government exceptions.
  • Ensure accountability in state surveillance.

Conclusion: India stands at a critical juncture in deploying AI-powered surveillance. While technological advancements promise enhanced governance and law enforcement, they must be balanced against constitutional rights. A proactive regulatory approach that is aligned with international best practices can ensure that AI serves the public interest without compromising civil liberties.