Context
- India faces a persistent challenge of low tax-GDP ratio (16.36%) and significant tax evasion (~4.3% revenue loss annually).
- To address this, the Income Tax Department launched Project Insight (2017; operational 2019), using AI and data analytics to improve tax compliance and governance.
Key Features of AI-based Tax Governance (Project Insight)
- Integrated data-driven system
- Uses multiple data sources: banking, GST, property, securities, credit cards, high-value transactions.
- Builds a 360-degree financial profile of taxpayers.
- Core components
- INTRAC (analytics engine): Detects inconsistencies using AI.
- Compliance Processing Centre: Ensures behavioural compliance.
- NUDGE strategy: Sends non-intrusive reminders (SMS/email) for voluntary correction.
- Focus areas
- Promote voluntary compliance.
- Reduce tax evasion and enforcement bias.
- Improve efficiency and taxpayer services.
Achievements and Emerging Trends
- Over 1 crore revised returns filed since 2020–21, generating ₹11,000 crore additional tax.
- 62% compliance in foreign income disclosure campaigns.
- Detection of ₹70,000 crore suppressed sales in restaurant sector using AI tools.
- Faster refunds: processing time reduced from 93 days to 17 days.
- Increasing global trend: countries like U.S., U.K., Australia, Italy using similar systems.
Advantages of AI in Tax Governance
- Improved compliance: Helps detect mismatches between income and spending patterns.
- Efficient enforcement: Enables risk-based targeting of high-value evasion cases.
- Administrative efficiency: Automates routine tasks, allowing officials to focus on complex cases.
- Better taxpayer services: Facilitates easier filing, quicker refunds, and AI-based assistance.
- Reduced human bias: Moves towards objective, data-driven decision-making.
Challenges and Concerns
- Data quality issues: AI depends on data accuracy; may generate false positives due to complex financial behaviour.
- Algorithmic bias: Risk of reinforcing existing socio-economic or regional biases.
- Lack of transparency: Limited explainability of AI decisions affects trust and accountability.
- Due process concerns: Taxpayers may face difficulty in challenging automated decisions.
- Privacy and security risks: Handling sensitive financial data increases risks of data breaches and misuse.
- Institutional gaps: Absence of an AI ombudsman, independent audits, and public reporting mechanisms.
Way Forward
- Ensure data integrity: Improve data quality and integration across systems.
- Strengthen accountability: Introduce AI audits, transparency norms, and public reporting of outcomes.
- Human oversight: Maintain human-in-the-loop systems for critical decisions.
- Protect taxpayer rights: Ensure clear grievance redressal and appeal mechanisms.
- Address bias and fairness: Regularly evaluate algorithms to prevent systemic discrimination.
- Institutional reforms: Establish an AI ombudsman and regulatory framework for ethical AI use.
Conclusion
- AI-driven tax governance offers a powerful tool to enhance compliance, efficiency, and revenue mobilisation. However, its success depends on balancing technological innovation with ethical safeguards, ensuring that the system remains fair, transparent, and trustworthy for citizens.
