AI-powered tax governance in India and its challenges

UPSC Syllabus: Gs Paper 3- Indian economy

Introduction

India faces a low tax-GDP ratio (16.36%) and high tax evasion (around 4.3% annual revenue loss). At the same time, AI is gaining importance in governance, as seen at the India AI Impact Summit (February 2026). In this situation, the use of AI in tax administration through Project Insight shows a shift towards data-driven governance to improve compliance, fairness, and revenue mobilisation.

AI in Tax Governance in India

1. Project Insight and core objective: Project Insight, launched in 2017 and operational in 2019, aims to increase voluntary compliance, reduce tax evasion, and ensure fair and unbiased tax enforcement through AI and data analytics.
2. Three components of Project Insight:

• INTRAC (analytical engine): INTRAC uses AI to analyse data from banking, property, securities, GST, credit cards, and high-value transactions to build a 360-degree financial profile of taxpayers.
• Compliance Management Centralized Processing Centre: It uses INTRAC data to monitor taxpayer behaviour and guide compliance actions.
• NUDGE strategy for voluntary compliance: It sends SMS and email reminders to help taxpayers correct or confirm their returns.

3. Working mechanism of AI system: AI identifies gaps between declared income and actual financial activities, enabling detection of potential tax evasion.

Benefits of AI in tax governance:

1. Risk-based prioritisation: AI helps authorities focus on high-risk and complex evasion cases, improving enforcement efficiency.
2. Automation of routine tasks: AI reduces manual work, allowing officials to focus on cases requiring human judgment.
3. Improved taxpayer services: AI supports correct return filing, chatbot assistance, and prevention of tax scams, improving compliance experience.

Effectiveness and Global Validation of AI in tax governance

1. Rise in voluntary compliance: Since 2020-21, over one crore revised returns have been filed, generating ₹11,000 crore additional tax revenue.
2. Improved foreign asset disclosure: In targeted campaigns, 62% taxpayers corrected returns, and 30,161 filers disclosed ₹29,208 crore assets and ₹1,089 crore income.
3. Correction of false deductions: NUDGE campaigns led to correction of ₹963 crore false claims and payment of ₹410 crore additional taxes.
4. Faster refund processing: Refund time reduced from 93 days to 17 days, improving efficiency and taxpayer experience.
5. Detection of large-scale evasion: AI tools uncovered ₹70,000 crore suppressed turnover in restaurants using advanced manipulation methods.
6. Global acceptance of AI systems: Countries like Australia, Italy, the UK, and the US have adopted similar AI-based systems and generated additional revenue.

Challenges and Risks in AI-Based Tax Governance

1. Data quality and false positives: AI depends on data quality and may wrongly flag legitimate financial behaviour as evasion, especially in complex income cases.
2. Algorithmic bias in decision-making: AI models may reflect existing socio-economic or geographic biases, leading to unequal targeting of taxpayers.
3. Lack of transparency and explainability: Taxpayers may not know why they are flagged or how decisions are made, weakening trust in the system.
4. Limited human oversight: Absence of easy human review forces taxpayers to justify flagged transactions, increasing compliance burden.
5. Privacy and data security risks: Use of sensitive financial data creates risk of misuse and cyber threats.
6. Institutional gaps in AI governance: Lack of AI ombudsperson, audit systems, and public reporting of errors reduces accountability.

What should be done?

1. Improve data quality systems: Strong data systems are needed to reduce false positives and incorrect targeting.
2. Ensure human-in-the-loop oversight: Human review must be included in key decisions to protect fairness and due process.
3. Enhance transparency and accountability: Taxpayers should know how AI decisions are made and how to challenge them.
4. Strengthen data protection frameworks: Robust safeguards are required to protect sensitive financial and personal data.
5. Establish institutional safeguards: Create systems for independent audits, ombudsperson, and reporting of error and appeal rates.
6. Maintain trust in tax system: Strong governance is needed to prevent AI systems from becoming opaque and surveillance-driven.

Conclusion

AI can strengthen tax compliance, improve efficiency, and increase revenue. However, weak safeguards can reduce fairness and trust. India must ensure transparency, accountability, human oversight, and data protection in AI systems. A balanced approach will help build a tax system that is both effective and trusted, ensuring sustainable compliance and fair governance.

Question for practice:

Examine how Artificial Intelligence is transforming tax governance in India, highlighting its effectiveness, key challenges, and the need for safeguards.

Source: The Hindu