[20th March 2026] The Hindu OpED: AI-powered tax governance in India and its challenges

Mentor’s Comment

The growing application of Artificial Intelligence in tax administration has significant implications for revenue mobilisation and governance in India. The Income Tax Department’s Project Insight (PI) represents a major shift towards data-driven tax administration. It leverages Artificial Intelligence and data analytics to enhance compliance, detect evasion, and improve revenue outcomes. 

What is Project Insight (PI) and how does it function?

1. Project Insight (PI): Establishes a data-driven tax intelligence system to strengthen compliance and enforcement.
2. Income Tax Transaction Analysis Centre (INTRAC): Processes financial data from banks, GST, property, and securities to generate taxpayer insights
3. 360-degree Profiling: Integrates multi-source financial data to build comprehensive taxpayer profiles
4. Non-intrusive Usage of Data to Guide and Enable (NUDGE) Strategy: Uses behavioural nudges such as SMS and emails to prompt voluntary compliance
5. Compliance Management Centralised Processing Centre: Ensures behavioural monitoring and correction of inaccurate filings

How does AI improve tax compliance and administrative efficiency?

1. Voluntary Compliance: Enables self-correction; over one crore revised returns filed since 2021
2. Targeted Enforcement: Identifies high-risk taxpayers; 19,501 individuals contacted under NUDGE campaign
3. Automation of Processes: Reduces routine workload; allows focus on complex assessments
4. Service Delivery: Assists taxpayers in filing returns and resolving queries through automated systems
5. Efficiency Gains: Reduces refund processing time from 93 days to 17 days

What are the measurable outcomes of AI-driven tax governance?

1. Revenue Augmentation: Generates ₹11,000 crore additional tax collection
2. Foreign Asset Disclosure: ₹1,089 crore declared under foreign income reporting
3. Digital Asset Tracking: ₹29,208 crore in overseas assets including cryptocurrencies identified
4. False Claim Correction: ₹963 crore corrected under NUDGE campaign
5. Additional Tax Payments: ₹410 crore realised from compliance actions
6. Evasion Detection: ₹70,000 crore suppressed turnover identified since 2019-20
7. Fraud Techniques Identified: Fake invoices, sales data manipulation, post-billing modifications

What are the challenges related to data quality and accuracy?

1. Data Dependence: Ensures outcomes depend on quality and completeness of input data
2. False Positives: Flags legitimate transactions (e.g., joint family structures, clerical errors) as suspicious
3. Error Propagation: Inaccurate data leads to flawed enforcement actions
4. Administrative Burden: Increases grievance redressal workload

How does algorithmic bias affect fairness in tax enforcement?

1. Historical Bias Replication: Uses past enforcement data, reinforcing socio-economic disparities
2. Geographical Skew: Targets specific regions or taxpayer categories disproportionately
3. International Example: Dutch childcare benefits scandal demonstrates risks of biased AI systems
4. Equity Concerns: Undermines fairness and trust in taxation

Why is explainability critical in AI-based tax systems?

1. Transparency Requirement: Ensures taxpayers understand reasons for scrutiny
2. Right to Appeal: Facilitates challenge to algorithmic decisions
3. Human Oversight: Maintains human-in-the-loop for high-impact decisions
4. Legal Validity: Supports principles of natural justice and due process

What are the concerns related to data privacy and security?

1. Sensitive Data Handling: Involves financial and personal taxpayer information
2. Cybersecurity Risks: Expands attack surface for data breaches
3. Surveillance Concerns: Enables potential misuse of taxpayer data
4. Regulatory Gaps: Highlights need for AI-specific safeguards

Why is institutional oversight necessary in AI governance?

1. AI Ombudsman Requirement: Establishes independent grievance redressal
2. Algorithm Audits: Ensures external verification of AI systems
3. Public Disclosure: Reports false positives and system accuracy
4. Trust Building: Enhances legitimacy of tax administration

Conclusion

AI-based tax governance improves compliance and revenue outcomes. However, risks related to bias, privacy, and accountability require institutional safeguards. A balance between efficiency and fairness remains essential.