INSURANCE DISCRIMINATION BASED ON PREDICTIONS: LEGAL FRAMEWORK, ETHICAL RESTRICTIONS AND JUDICIAL INTERPRETATION

1. INTRODUCTION

Insurance discrimination based on predictions refers to the unfair or unjustified treatment of individuals by insurance companies using predictive information, statistical models, artificial intelligence, genetic data, medical histories, behavioural patterns or algorithmic assessments to estimate future risks.

Modern insurance companies increasingly use predictive technologies to evaluate the probability of illness, disability, accidents, mortality and financial loss. Although risk assessment is a fundamental component of insurance, discriminatory practices may arise when insurers rely on inaccurate predictions, sensitive personal information, historical inequalities or characteristics that individuals cannot reasonably control.

For example, an insurance company may increase premiums, deny coverage, impose restrictive conditions or refuse renewal because an algorithm predicts that an individual is likely to develop a particular disease in the future.

Such practices raise important legal and ethical questions concerning equality, privacy, consumer protection, transparency, genetic discrimination and fairness in financial services.

The central objective of insurance regulation is to distinguish legitimate actuarial risk assessment from arbitrary, misleading or discriminatory predictive decision-making.

2. FUNDAMENTAL ETHICAL PRINCIPLES

A. PRINCIPLE OF NON-DISCRIMINATION

Insurance companies should not unfairly disadvantage individuals because of genetic characteristics, disability, gender, ethnicity, socioeconomic background or other protected characteristics.

Risk classification must have a legitimate actuarial or underwriting justification and comply with applicable anti-discrimination laws.

B. PRINCIPLE OF FAIRNESS

Predictive models must produce reasonably accurate and proportionate assessments. Individuals should not suffer excessive financial penalties because of unreliable predictions or assumptions unsupported by relevant evidence.

C. PRINCIPLE OF PRIVACY

Insurance companies must respect the confidentiality of personal, medical, genetic and behavioural information.

Collection and use of sensitive information must comply with applicable data-protection and insurance regulations.

D. PRINCIPLE OF TRANSPARENCY

Insurers should provide appropriate explanations of material underwriting decisions, including premium increases, coverage exclusions and rejection of applications, subject to applicable legal requirements.

E. PRINCIPLE OF HUMAN ACCOUNTABILITY

Automated predictive systems should not eliminate institutional responsibility. Insurers must maintain appropriate oversight of algorithmic decisions and provide mechanisms for reviewing potentially unfair outcomes.

F. PRINCIPLE OF SOCIAL JUSTICE

Insurance should provide meaningful financial protection rather than systematically excluding individuals with greater health or economic vulnerabilities.

3. LEGAL AND REGULATORY FRAMEWORK IN INDIA

A. INSURANCE ACT, 1938

The Insurance Act establishes the principal statutory framework governing insurance business in India.

It regulates insurers, insurance operations and relevant obligations concerning policy administration.

Insurance companies must conduct underwriting and policy-related activities consistently with applicable statutory and regulatory requirements.

B. INSURANCE REGULATORY AND DEVELOPMENT AUTHORITY OF INDIA ACT, 1999

The IRDAI Act establishes the Insurance Regulatory and Development Authority of India.

IRDAI supervises insurance-sector development, policyholder protection and regulatory compliance.

C. IRDAI (PROTECTION OF POLICYHOLDERS' INTERESTS, OPERATIONS AND ALLIED MATTERS OF INSURERS) REGULATIONS, 2024

These regulations establish requirements concerning policyholder protection, disclosures, servicing and insurance-related obligations.

Insurers must provide appropriate information and comply with applicable standards governing policyholder treatment.

D. CONSTITUTION OF INDIA

Article 14 guarantees equality before the law and equal protection of the laws.

Article 21 protects life and personal liberty, including privacy.

These constitutional protections are particularly relevant to discriminatory state action and may also influence the regulation of private insurance activities through applicable legal principles.

E. CONSUMER PROTECTION ACT, 2019

Insurance services fall within the consumer protection framework.

Unfair trade practices, deficient services and misleading representations may create consumer liability depending on the circumstances.

F. DIGITAL PERSONAL DATA PROTECTION ACT, 2023

Where insurers process digital personal data for predictive underwriting, applicable data-protection obligations must be considered, subject to the Act's commencement, rules and exemptions.

4. IMPORTANT JUDICIAL DECISIONS

There is no established Indian Supreme Court judgment specifically deciding algorithmic insurance discrimination based on future predictions. However, the following decisions provide important direct or analogous legal principles.

1. UNITED INDIA INSURANCE CO. LTD. v. MANUBHAI DHARMASINHBHAI GAJERA (2008) — DIRECT ANALOGICAL INSURANCE AUTHORITY

The Supreme Court considered insurance-related contractual obligations and the principles governing renewal and policyholder interests.

Relevance: Insurance companies must exercise their contractual and regulatory powers consistently with applicable legal requirements. Predictive assessments cannot automatically justify arbitrary treatment contrary to policy terms or governing regulations.

2. LIFE INSURANCE CORPORATION OF INDIA v. CONSUMER EDUCATION & RESEARCH CENTRE (1995) — DIRECT INSURANCE FAIRNESS AUTHORITY

The Supreme Court examined discriminatory insurance policy conditions and the obligations of a public-sector insurer.

Relevance: The judgment is important for understanding fairness, equality and non-arbitrary treatment in insurance practices. It provides a useful foundation for examining discriminatory underwriting practices.

3. NATIONAL INSURANCE CO. LTD. v. HARSHA MEHRA (2003) — ANALOGICAL INSURANCE AUTHORITY

The Supreme Court considered issues concerning insurance policy interpretation and contractual obligations.

Relevance: Insurers must comply with applicable policy conditions and cannot disregard contractual responsibilities through unsupported underwriting assumptions.

4. GENERAL ASSURANCE SOCIETY LTD. v. CHANDMULL JAIN (1966) — DIRECT INSURANCE CONTRACT AUTHORITY

The Supreme Court examined the legal principles governing insurance contracts and the importance of interpreting insurance terms according to their contractual meaning.

Relevance: Predictive underwriting decisions must be assessed against the actual contractual terms, applicable disclosure obligations and relevant insurance regulations.

5. K.S. PUTTASWAMY v. UNION OF INDIA (2017) — CONSTITUTIONAL PRIVACY AUTHORITY

The Supreme Court recognised privacy as a fundamental right under Article 21.

Relevance: The collection and use of genetic, medical and behavioural information for predictive insurance assessments must respect applicable privacy protections.

6. ANUJ GARG v. HOTEL ASSOCIATION OF INDIA (2008) — DIRECT ANALOGICAL ANTI-DISCRIMINATION AUTHORITY

The Supreme Court examined discriminatory restrictions based on gender and emphasised substantive equality and constitutional protection against stereotypes.

Relevance: Insurance companies should not rely on unsupported generalisations or discriminatory assumptions when predicting the future risks of individuals.

7. JEEJA GHOSH v. UNION OF INDIA (2016) — DISABILITY AND DIGNITY AUTHORITY

The Supreme Court emphasised dignity, equality and the rights of persons with disabilities.

Relevance: Predictive insurance systems should not automatically disadvantage individuals because of disability or assumptions about their future capabilities.

5. MAJOR FORMS OF PREDICTIVE INSURANCE DISCRIMINATION

A. GENETIC PREDICTION DISCRIMINATION

Insurers may attempt to use genetic information to estimate an individual's likelihood of developing inherited diseases.

Ethical concerns arise when individuals are penalised for genetic characteristics that they cannot control, particularly where predictions are uncertain or medically inconclusive.

B. MEDICAL RISK PREDICTION

Predictive models may use medical histories, laboratory reports and health information to determine future insurance risks.

Such assessments must respect applicable legal requirements and avoid unsupported medical assumptions.

C. BEHAVIOURAL PREDICTION

Insurance companies may use behavioural information, lifestyle indicators or digital activity to estimate future risks.

Unfairness may arise when inaccurate behavioural profiles or intrusive data collection influence premiums or coverage.

D. AGE-BASED PREDICTION

Age may be relevant to legitimate actuarial assessments, but its use must comply with applicable insurance regulations and cannot justify prohibited discrimination.

E. DISABILITY-BASED PREDICTION

Individuals with disabilities may experience unfair exclusions or higher premiums because of assumptions about future medical expenses.

Insurers must comply with applicable disability rights and insurance regulations.

F. AI-BASED PREDICTIVE DISCRIMINATION

Artificial intelligence systems may reproduce historical inequalities or generate inaccurate risk classifications.

Insurers should evaluate model accuracy, data quality, discriminatory effects and decision-making accountability.

6. MAJOR LEGAL AND ETHICAL RESTRICTIONS

A. RESTRICTIONS ON UNJUSTIFIED DATA COLLECTION

Insurers should collect and process only information supported by a legitimate and lawful purpose.

B. PROHIBITION OF MISLEADING UNDERWRITING PRACTICES

Insurance companies should not present uncertain algorithmic predictions as established medical or actuarial facts.

C. PROTECTION AGAINST ARBITRARY PREMIUM INCREASES

Premium decisions must comply with applicable regulatory requirements, contractual conditions and legitimate underwriting standards.

D. TRANSPARENCY IN AUTOMATED DECISION-MAKING

Policyholders should have appropriate access to information concerning material adverse underwriting decisions and available grievance mechanisms.

E. PROTECTION OF GENETIC INFORMATION

Genetic information requires particularly careful treatment because it may reveal information about both individuals and their biological relatives.

F. HUMAN REVIEW AND CORRECTION

Insurers should establish procedures for correcting inaccurate information and reviewing potentially discriminatory decisions.

7. INTERNATIONAL LEGAL AND ETHICAL FRAMEWORK

European Union General Data Protection Regulation (GDPR), 2016: Establishes requirements concerning personal data processing, sensitive data and certain automated individual decision-making.

EU Gender Directive 2004/113/EC: Addresses equal treatment between men and women in access to and supply of goods and services, including insurance, subject to its provisions and recognised exceptions.

United States Genetic Information Nondiscrimination Act (GINA), 2008: Restricts the use of genetic information in health insurance and employment, but generally does not extend its protections to life, disability or long-term care insurance.

Universal Declaration of Human Rights, 1948: Articles 2 and 7 provide important principles concerning non-discrimination and equality.

International Covenant on Civil and Political Rights, 1966: Provides protections concerning privacy and equality under Articles 17 and 26.

These instruments differ in geographical applicability and legal enforceability.

8. PRACTICAL COMPLIANCE REQUIREMENTS

Insurance companies, underwriting departments and technology providers should:

Establish transparent underwriting policies.

Ensure predictive models are based on relevant and reliable information.

Conduct regular algorithmic fairness assessments.

Avoid prohibited discriminatory classifications.

Protect sensitive medical and genetic information.

Maintain appropriate data security arrangements.

Provide policyholders with relevant information about adverse decisions.

Establish procedures for correcting inaccurate personal information.

Maintain human oversight of significant underwriting decisions.

Comply with applicable IRDAI regulations.

Provide effective grievance-redressal mechanisms.

Maintain records supporting material premium and coverage decisions.

9. LIABILITY AND REMEDIES

Unfair predictive insurance practices may result in:

Complaints before the insurer's grievance-redressal mechanism.

Complaints before the Insurance Ombudsman, where jurisdictional requirements are satisfied.

Regulatory action by IRDAI.

Consumer commission proceedings.

Contractual remedies for wrongful repudiation or breach.

Data-protection enforcement where applicable.

Constitutional remedies against unlawful state action.

The availability of a remedy depends on the nature of the discrimination, the insurance product, the applicable regulations and the evidence supporting the claim.

10. CONCLUSION

Insurance discrimination based on predictions presents an important challenge to modern insurance governance. Predictive analytics can improve underwriting accuracy and financial risk management, but their use must not undermine equality, privacy, human dignity and consumer protection.

Indian constitutional principles, insurance legislation, IRDAI regulations and consumer protection laws provide safeguards against arbitrary and unfair insurance practices.

Judicial decisions such as LIC v. Consumer Education & Research Centre, Puttaswamy, Anuj Garg and Jeeja Ghosh establish important principles concerning insurance fairness, privacy, equality and protection against discrimination.

Ultimately, legitimate insurance risk assessment must be distinguished from unjustified predictive discrimination. Insurance companies should use reliable evidence, transparent methodologies, proportionate underwriting criteria and effective human oversight. Predictive technology must support fair insurance services rather than become a mechanism for excluding individuals on the basis of uncertain future possibilities.