Key Terms & Concepts — UPSC Mains
Predictive Policing
"The use of AI, machine learning, and data analytics to forecast where crimes are likely to occur or identify individuals likely to commit crimes, allowing pre-emptive police action"
Predictive policing is a law enforcement strategy that uses artificial intelligence, big data analytics, and statistical modelling to forecast crime patterns, predicting where crimes are likely to occur (place-based prediction) or identifying individuals at high risk of committing or being victimised by crime (person-based prediction). Outputs are used to allocate police resources, increase patrols in flagged areas, or flag individuals for proactive investigation. India's DRDO developed 'Prajna', an AI-based satellite imagery and surveillance system used by MHA for monitoring sensitive areas, representing India's entry into AI-enabled security intelligence.
Relevant for GS3 (Internal Security, policing technology, surveillance) and GS2 (Governance, civil liberties, privacy, accountability). Predictive policing raises fundamental tension between crime prevention efficiency and civil liberties, particularly the right to privacy (Justice K.S. Puttaswamy verdict, 2017) and the presumption of innocence. UPSC essays and Ethics papers frequently examine technology-governance tensions.
- 1 Types, place-based (hotspot mapping) and person-based (risk scoring of individuals or groups)
- 2 Examples, PredPol (USA), Chicago's Strategic Subject List, UK's National Data Analytics Solution
- 3 India, DRDO's Prajna system (2026); AI-based satellite imagery analysis for MHA; Crime and Criminal Tracking Network (CCTNS)
- 4 Criticism, algorithmic bias perpetuates racial/caste profiling; surveillance creep; lacks transparency
- 5 Legal concerns, pre-crime intervention undermines presumption of innocence; no statutory framework in India
- 6 Puttaswamy judgment (2017), right to privacy as fundamental right; predictive surveillance must meet proportionality test
- 7 PDPB/DPDP Act 2023, does not specifically regulate law enforcement data use; creates oversight gap
- 8 Accountability gap, AI decisions in policing lack audit trails, grievance mechanisms, or judicial oversight
- 9 Global context, EU AI Act classifies real-time biometric surveillance in public spaces as 'unacceptable risk'
DRDO's Prajna system, deployed by MHA for surveillance in sensitive border areas and conflict zones, uses AI-based analysis of satellite imagery to detect unusual movement patterns, a form of predictive intelligence that raises questions about oversight, data retention, and the legal framework governing its use against civilians.
