Public-Health Predictive Policing: Legality

Introduction

Public-health predictive policing refers to the use of data analytics, artificial intelligence (AI) and statistical models to predict health-related risks and guide government or law-enforcement interventions. It may involve predicting disease outbreaks, identifying locations with higher infection rates, forecasting non-compliance with quarantine orders, or using health data to identify individuals considered likely to violate public-health regulations.

Predictive systems can help authorities allocate medical resources and prevent disease transmission. However, when used to monitor, profile or penalise individuals, they raise serious concerns about privacy, discrimination, due process, accuracy and the misuse of health information.

The central legal question is whether predictive public-health interventions are necessary, evidence-based, proportionate and supported by law—or whether they amount to arbitrary surveillance and discriminatory treatment.

Legal Framework in India

The Constitution of India provides important safeguards:

Article 14: Protects equality before the law and prohibits arbitrary State action.

Article 19: Protects specified freedoms, including movement and expression, subject to constitutionally permitted restrictions.

Article 21: Protects life and personal liberty, including privacy, dignity and procedural fairness.

The Supreme Court’s recognition of privacy as a fundamental right means that public-health data collection and predictive surveillance must be justified by a legitimate purpose and appropriate safeguards. A government cannot rely solely on technological capability to justify intrusive monitoring.

Public-health predictive systems may also involve the Digital Personal Data Protection Act, 2023, subject to its commencement, applicable rules and exemptions. Where personal data is processed, the legal basis, purpose, security and applicable rights must be considered.

Case Laws

1. Justice K.S. Puttaswamy (Retd.) v. Union of India, (2017) 10 SCC 1

The Supreme Court recognised privacy as a fundamental right under Articles 14, 19 and 21. It held that State interference with privacy must satisfy legality, a legitimate State aim and proportionality. The judgment also recognised informational privacy and the risks of profiling through the convergence of databases.

This is a central authority for predictive policing: public-health authorities cannot collect and combine sensitive personal data merely because technology makes it possible.

Indian Kanoon+1

2. Anuradha Bhasin v. Union of India, (2020) 3 SCC 637

The Supreme Court held that restrictions affecting fundamental rights must be supported by law, pursue a legitimate aim and be proportionate. It also emphasised the need for publication and review of orders restricting access to the internet.

Although the case concerned internet restrictions, its proportionality framework applies by analogy to predictive public-health surveillance, especially where the system affects movement, access to services or expression.

Indian Kanoon

3. Malak Singh v. State of Punjab & Haryana, (1981) 1 SCC 420

The Supreme Court considered police surveillance and held that surveillance must remain within legal limits. Police cannot include individuals in surveillance registers arbitrarily, nor may surveillance intrude so extensively as to impair fundamental freedoms or offend individual dignity.

This case is relevant to predictive policing because an algorithmic risk score should not itself become a basis for intrusive surveillance without lawful grounds and safeguards.

Indian Kanoon

Legal Safeguards and Limits

A predictive public-health policing system should satisfy the following requirements:

Clear legal authority: The government must identify the law authorising data collection, analysis and any resulting intervention.

Legitimate public-health objective: The system must address a genuine health risk, not political dissent or unrelated law-enforcement goals.

Necessity and proportionality: Less intrusive measures, such as anonymous outbreak mapping or voluntary testing, should be considered first.

Accuracy and bias review: Models should be tested for false positives, discriminatory effects and unequal impact on vulnerable communities.

Human oversight: A predictive score should not automatically trigger detention, punishment, denial of healthcare or other serious consequences.

Notice and remedy: Individuals should have a meaningful way to challenge inaccurate data or adverse decisions, subject to lawful and narrowly tailored exceptions.

Data minimisation and security: Health data should be collected only as necessary, retained for a defined period and protected against unauthorised access.

Conclusion

Public-health predictive policing may be lawful when it is genuinely directed toward preventing disease or protecting public safety and operates under clear legal authority. However, predictive algorithms cannot replace constitutional safeguards. The principles in Puttaswamy, Anuradha Bhasin and Malak Singh require legality, proportionality, non-arbitrariness and respect for dignity. The use of predictive systems should therefore focus on public-health support and resource planning, not automatic punishment or indiscriminate surveillance of individuals.