Facial Recognition by Police 

Facial recognition by police refers to the use of AI-based biometric technology to identify or verify individuals by analyzing facial features from images or live video feeds. It is increasingly used for:

• Crime investigation
• Surveillance in public spaces
• Missing persons detection
• Crowd monitoring

However, it raises serious constitutional concerns related to privacy, surveillance, accuracy, and misuse of power.

1. Meaning and Scope

Facial recognition systems (FRS) work by:

• Capturing facial image (CCTV / mobile / database)
• Extracting biometric facial features
• Matching with stored databases
• Generating identity probability match

👉 Police use it for:

• Suspect identification
• Tracking criminal movements
• Real-time surveillance

2. Constitutional Basis (India)

(a) Article 21 – Right to Life & Privacy

• Protects personal autonomy and dignity

(b) Article 14 – Equality

• Prevents arbitrary or biased surveillance

(c) Article 19 – Freedom of movement and expression

• Excessive surveillance creates chilling effect

3. Legal Framework in India

There is no specific facial recognition law, but regulation comes from:

• Information Technology Act, 2000
• Criminal Procedure Code (CrPC)
• Aadhaar Act (biometric principles)
• Police guidelines and executive orders

👉 Problem: Lack of clear statutory safeguards

4. Key Legal Concerns

(1) Privacy Violation

• Constant surveillance without consent

(2) Accuracy Issues

• False positives (wrong identification)
• Bias against certain groups

(3) Mass Surveillance Risk

• Monitoring entire populations

(4) Lack of Transparency

• Secret databases and algorithms

5. Proportionality Test (Core Legal Standard)

Any facial recognition use must pass:

1. Legitimate aim (crime prevention)
2. Necessity (no less intrusive method)
3. Proportionality (limited and targeted use)
4. Safeguards (oversight + accountability)

6. Important Case Laws

(1) Justice K.S. Puttaswamy v. Union of India (2017)

• Held: Privacy is a fundamental right under Article 21
• Relevance: Facial recognition is a form of biometric surveillance and must meet strict proportionality

(2) Kharak Singh v. State of Uttar Pradesh (1962)

• Held: Unauthorized surveillance violates personal liberty
• Relevance: Continuous facial tracking = modern form of domiciliary surveillance

(3) People’s Union for Civil Liberties v. Union of India (1997)

• Held: Telephone tapping requires strict safeguards
• Relevance: Extends to modern digital surveillance like facial recognition

(4) Gobind v. State of Madhya Pradesh (1975)

• Held: Privacy can be restricted only with compelling state interest
• Relevance: Police FRS must show strong justification

(5) Selvi v. State of Karnataka (2010)

• Held: Forced biometric tests violate Article 20(3) and privacy
• Relevance: Raises concerns about compulsory facial data collection

(6) Shreya Singhal v. Union of India (2015)

• Held: Restrictions on expression must be narrow and precise
• Relevance: Overbroad surveillance chills free expression and movement

(7) Anuradha Bhasin v. Union of India (2020)

• Held: Restrictions on fundamental rights must be temporary, necessary, and reviewable
• Relevance: Facial recognition deployment must be time-bound and reviewable

7. International Standards

(a) European Union (GDPR)

• Biometric data = sensitive personal data
• Strict consent and limitation rules

(b) United States

• Mixed system:
  • Some states ban police facial recognition
  • Requires warrants in certain cases

(c) UN Human Rights Principles

• Surveillance must be:
  • Legal
  • Necessary
  • Proportionate

8. Police Use: Permitted vs Restricted

✔ Permitted (limited conditions)

• Identifying missing persons
• Investigating serious crimes
• Court-authorized surveillance

❌ Not Permitted / Risky

• Mass surveillance of public without cause
• Real-time tracking of all citizens
• Secret databases without oversight
• Indiscriminate scanning in public spaces

9. Procedural Safeguards Needed

• Judicial authorization for deployment
• Data minimization rules
• Audit of algorithm accuracy
• Transparency reports
• Time-limited usage
• Independent oversight body

10. Key Legal Principles

1. Facial recognition = biometric surveillance → privacy sensitive
2. Must pass strict proportionality test
3. Requires legal authorization and safeguards
4. Mass surveillance is constitutionally suspect
5. Accuracy and bias must be addressed

11. Conclusion

Facial recognition by police sits at the intersection of security and civil liberties. While it can enhance crime detection and public safety, Indian constitutional law—especially after the Puttaswamy judgment—requires that such surveillance be strictly necessary, proportionate, and legally regulated. Without strong safeguards, it risks becoming a tool of mass surveillance and rights violation rather than targeted policing.