[Answered] While AI-powered tax governance in India promises enhanced compliance and efficiency, it transitions the administration into a realm of complex algorithmic risks. Discuss the outcomes and evaluate the operational legal challenges involved.

Introduction

Economic Survey 2025-26 notes India’s low tax-GDP ratio (~16.3%) and rising AI adoption. Budget 2026-27 prioritises digital governance, while NITI Aayog highlights AI-driven compliance as key to widening the tax base.

AI-Powered Tax Governance in India

AI and ML to revolutionizing tax administration, leading to significant increases in tax collection, enhanced detection of evasion, and improved compliance. Through initiatives like Project Insight and ADVAIT, the CBDT and the CBIC have implemented AI-driven tools that analyze vast amounts of data to identify discrepancies and target high-risk taxpayers.

Key AI Initiatives and Results

1. Revenue Generation: NUDGE campaigns have prompted over 1 crore updated returns and generated over ₹11,000 crore in tax revenue between 2021-25.
2. Asset & Evasion Detection: AI identified over ₹29,000 crore in undisclosed foreign assets (including crypto) and helped detect ₹30,000 crore in GST evasion in FY 2023-24 via ADVAIT.
3. Systemic Tools: Project Insight provides a 360-degree taxpayer profile, while CASS enables AI-driven, rapid scrutiny of high-risk returns.
4. Compliance & Verification: AI-driven audits (like on Section 80GGC) and automated notices based on mismatch analysis have significantly improved voluntary compliance and tax recovery.
5. Digital Integration: The upcoming PAN 2.0 project will feature advanced AI fraud detection, moving towards a data-driven, proactive compliance model.

Operational Challenges

1. Data Quality and Integration: Fragmented legacy systems and inconsistent taxpayer data across GSTN, banks, and MCA-21 create incomplete 360° profiles, leading to false positives/negatives in risk flagging. For Example-Mismatch between informal sector transactions and reported income may wrongly flag taxpayers.
2. Algorithmic Errors and Tax Terrorism: AI models lack transparency (“black-box problem”). Difficulty in explaining why a taxpayer is flagged. Reduces trust in tax administration and complicates dispute resolution.
3. Scalability & Infrastructure: Processing billions of transactions requires massive compute power; IndiaAI Mission’s 38,000+ GPUs help, but real-time analytics for 70 crore+ taxpayers remain constrained.
4. Over-Reliance on Automation: Excessive dependence may reduce human discretion in nuanced cases. Automated decisions may ignore contextual realities. For Example-genuine income fluctuations.

Legal Challenges

The framework operates in a grey zone:

1. Article 14 Violation: Unequal treatment arises when algorithms flag similar transactions differently based on opaque logic.
2. Article 21 Intrusion: Mass financial surveillance without proportionality or judicial oversight encroaches on privacy (post-Puttaswamy).
3. Section 69A Parallel: Broad blocking powers under IT Act mirror tax nudges; lack of pre-decisional hearing and reasoned orders breaches natural justice.
4. Data Protection Gaps: DPDP Act 2023 applies, but enforcement lags; no specific safeguards exist for tax AI processing sensitive financial data.

Broader Economic and Governance Implications

1. Fiscal Strengthening: Improved compliance helps raise tax-GDP ratio, enabling welfare spending.
2. Formalisation of Economy: AI-driven tracking incentivises shift from informal to formal sector.
3. Global Alignment: Aligns with OECD trends on digital tax administration.

Way Forward

1. Mandate Algorithmic Impact Assessments (AIAs) with public disclosure for high-risk tax AI systems.
2. Establish independent AI Ethics Oversight Committee under CBDT with judicial and civil-society representation.
3. Implement explainable AI (XAI) models with taxpayer-facing reasons summaries for nudges/flags.
4. Create grievance redressal tribunals for algorithmic decisions with fast-track appeal.
5. Integrate continuous bias audits and diverse training data to prevent disproportionate impact.

Conclusion

Technology must empower citizens, not diminish rights; like Economic Survey insights, AI-driven taxation must balance efficiency with transparency, accountability, and constitutional safeguards to sustain democratic trust.