Read Time: 12 min

Direct Taxation

February 2026

Artificial Intelligence in Income Tax Administration Friend or Foe.

Introduction

Picture of by Anjali Darak
by Anjali Darak

Manager - Direct Tax

Artificial Intelligence (AI) is rapidly transforming Income Tax administration across the globe. In India, AI-driven systems now influence return processing, scrutiny selection, faceless assessments and enforcement actions. While the objective is enhanced efficiency, transparency and revenue protection, the growing reliance on algorithms raises critical questions around taxpayer rights, accountability and fairness.

This leads us to the core debate: Is AI a friend improving tax governance or a foe undermining natural justice?

Coming to this month’s, Taxation Times, here’s what we have:

  1. An article on “Artificial Intelligence in Income Tax Administration: Friend or Foe?”.
  2. Case Laws from various courts & jurisdictions.
  3. Tax Compliance Calendar – February 2026;
  4. Circulars & Notifications – January 2026;
  5. Tax News from around the world

We hope that you find this month’s edition of the Taxation Times useful. In case you have any feedback or need us to include any information to make this issue more informative, please feel free to write to us at info@uja.in.

Happy New Year!
Happy Reading!

Best Regards,
UJA Tax Team

Artificial Intelligence in Income Tax Administration: Friend or Foe?

Abstract

The Income Tax administration in India has undergone a significant transformation with the adoption of Artificial Intelligence (AI) and data analytics. From return processing and scrutiny selection to faceless assessments and enforcement actions, AI-driven systems are increasingly shaping the taxpayer’s experience. While the stated objectives are efficiency, transparency and reduction of discretion, the growing reliance on algorithmic decision-making raises fundamental concerns relating to natural justice, accountability and fairness. This article critically examines whether AI in Income Tax administration is a friend, facilitating better governance or a foe, creating new challenges for taxpayers and professionals.

Introduction

Digitalisation of tax administration is no longer optional—it is inevitable. In recent years, the Indian Income Tax Department has aggressively leveraged technology, culminating in the introduction of faceless assessments, automated return processing and integrated reporting systems such as AIS and TIS. At the heart of this transformation lies Artificial Intelligence, which enables the processing and analysis of massive volumes of data at unprecedented speed.

However, taxation is not merely a computational exercise. It involves the interpretation of law, the appreciation of facts and the application of judgment. This dichotomy between algorithmic efficiency and human discretion forms the crux of the debate: can AI enhance tax administration without undermining taxpayer rights?

Role of AI in Income Tax Administration

AI is currently deployed across multiple stages of tax administration:

  • Return Processing: Automated processing of returns and issuance of intimations under section 143(1)
  • Scrutiny Selection: Risk-based selection using data analytics
  • Faceless Assessments: Allocation of cases, issue of notices and draft orders through system-driven workflows
  • Evasion Detection: Identification of high-risk transactions, shell entities and suspicious patterns
  • Data Integration: Cross-verification using third-party data from banks, GST systems, registrars and financial institutions

The scale and complexity of modern economic transactions make manual administration impractical, thereby justifying AI intervention.

AI as a Friend: Benefits to Tax Administration and Taxpayers

Efficiency and Speed

AI has significantly reduced processing time by automating repetitive and rule-based tasks. Refunds are issued faster, returns are processed swiftly and compliance monitoring is continuous rather than episodic.

Objective Scrutiny Selection

Traditional scrutiny selection was often perceived as arbitrary. AI enables:

  • Risk-based selection
  • Targeted scrutiny of high-risk cases
  • Reduced intrusion for compliant taxpayers

This marks a shift from subjective discretion to data-driven decision-making.

Reduction in Human Bias and Corruption

Faceless and AI-assisted systems limit direct interaction between taxpayers and officers, reducing:

  • Scope for coercion or undue influence
  • Regional and individual biases
  • Inconsistent treatment of similar cases

Improved Detection of Tax Evasion

AI excels at pattern recognition and anomaly detection. It can identify:

  • Circular transactions
  • Accommodation entries
  • Bogus losses
  • Benami arrangements

Such capabilities strengthen the Department’s enforcement machinery.

AI as a Foe: Challenges and Risks

Despite its advantages, AI introduces several systemic concerns.

Lack of Algorithmic Transparency

Taxpayers are rarely informed:

  • Why a case was selected for scrutiny
  • Which risk parameters were triggered
  • How third-party data was evaluated

This “black box” approach conflicts with transparency and weakens taxpayer confidence.

Mechanical and Overbroad Assessments

AI-generated notices often:

  • Flag mere mismatches without context
  • Ignore commercial realities and accounting principles
  • Treat timing differences as concealment

Taxation, however, requires a nuanced appreciation of facts—something algorithms struggle to replicate.

Natural Justice Concerns

Principles of natural justice require:

  • Opportunity to be heard
  • Reasoned orders
  • Application of mind

Automated workflows risk reducing assessments to box-ticking exercises, undermining these principles. Courts have repeatedly emphasized that technology cannot replace judicial or quasi-judicial reasoning.

Data Quality and False Positives

AI’s output is only as good as its input. Issues include:

  • Incorrect third-party reporting
  • Duplicate or outdated information
  • Incomplete transaction details

This leads to unnecessary litigation and compliance burden.

Accountability and Responsibility

A critical unanswered question remains:
Who is accountable for an AI-driven error?

    • The Assessing Officer?
    • The system designer?
    • The Department?

The absence of a clear accountability framework complicates redressal mechanisms.

Judicial Response and Legal Position

The Indian judiciary has consistently held that:

  • Technology is an enabling tool, not a substitute for human judgment
  • Faceless and AI-driven assessments must comply with statutory safeguards
  • Violation of natural justice renders assessments invalid

Courts have intervened where:

  • Adequate opportunity was not provided
  • Orders lacked reasoning
  • Automated processes led to unjust outcomes

The judicial trend clearly favours AI with human oversight, not autonomous decision-making.

Impact on Tax Professionals

For tax professionals, AI presents both challenges and opportunities.

Challenges

  • Increased compliance scrutiny
  • Need for continuous data reconciliation
  • Responding to system-generated notices

Opportunities

  • Shift from routine compliance to advisory roles
  • Focus on litigation strategy and risk management
  • Greater emphasis on documentation and tax governance

The role of the tax professional is evolving from a compliance executor to a strategic advisor.

Way Forward: Striking the Right Balance

For the Tax Administration

  • Develop explainable and auditable AI systems
  • Ensure meaningful human intervention at critical stages
  • Issue transparent guidelines on AI usage

For Policymakers

  • Define legal boundaries for algorithmic decision-making
  • Establish accountability frameworks
  • Balance revenue interests with taxpayer rights

For Taxpayers

  • Maintain robust documentation
  • Regularly reconcile reported data
  • Adopt proactive compliance strategies

Conclusion

Artificial Intelligence in Income Tax administration is neither inherently beneficial nor inherently harmful. Its effectiveness depends on responsible design, transparent deployment and strong human oversight.

When AI is used as an intelligent assistant, it can enhance efficiency, reduce corruption and improve compliance. When treated as an unquestionable authority, it risks eroding fairness, accountability and trust.

The future of tax administration lies not in replacing human judgment with machines, but in augmenting human decision-making through technology.

Case Laws

JANUARY, 2026
[[2026] 182 taxmann.com 296 (Ahmedabad - Trib.) IN THE ITAT AHMEDABAD BENCH SKZ Developers LLP v. ACIT/DCIT
JANUARY 9, 2026
Section 69A of the Income-tax Act, 1961 - Unexplained money (On-money)

Fact I :

  • A search under section 132 was conducted in the B-Safal & City Estate Group, during which loose slips and digital material indicating alleged cash transactions for the sale of units were found. The Assessing Officer held that the assessee-LLP formed part of the same group and concluded that it had received on-money on the sale of units in its projects ‘Privilon’ and ‘Paarijat Eclate’.
  • Based on seized material, the Assessing Officer estimated project cost at Rs. 399.31 crore, worked out construction costs at Rs. 5,000 and Rs. 4,600 per sq. ft. and noted that several registered sale deeds reflected lower basic rates. Concluding that no builder would sell below cost, he treated the difference as unaccounted-on money and made additions aggregating to Rs. 53.36 crore.
  • On appeal, the Commissioner (Appeals) held that the Assessing Officer erred in treating loose Excel sheets and working papers as conclusive proof without corroboration. He recorded that no cash trail, buyer confirmation, or utilization of alleged cash was found, and that the Assessing Officer acted on assumptions without independent verification.
  • However, noting seized cost-working files showing construction costs of Rs. 5,000 and Rs. 4,600 per sq. ft. and certain broker materials and WhatsApp chats showing negotiated rates of Rs. 6,100–7,350 per sq. ft. the Commissioner (Appeals) held that sale values could not rationally be below cost. He therefore estimated fair sale value at Rs. 6,500 per sq. ft., applied this uniformly (subject to higher registered sale values), and directed that only the embedded profit element be taxed at 17 percent.
  • On appeal by assessee before the Tribunal:

Held I:

  • The documents relied upon by the Assessing Officer were nothing but dumb documents as these loose sheets did not contain details of buyers, date of receipt of alleged on money, signature of buyer or agreement to sale. Therefore, the contention of the assessee that these loose papers were just quotations/offer price only cannot be brushed aside. Moreover, as noted by the Assessing Officer, in the relevant column of these seized documents, it was mentioned a ‘Offer Rate’. It is agreed with the contentions of the assessee that in the real estate business, offer rates are always on the higher side and are subject to negotiation. There is no evidence in the file that any transaction actually materialized at the rates in the relevant assessment year as depicted in the alleged seized documents. The assessee has also explained that the Assessing Officer has wrongly assumed/interpreted the various short words/abbreviations to his convenience and linked the same to assume higher sale rates of the units. He has also demonstrated that the Assessing Officer found notes saying ‘1P=’4L” and assumed this meant ‘1 Parking Spot = Rs. 4 Lakhs in cash. ‘The assessee, however, explained that ‘P’ actually stood for ‘Painting’ (interior finishing works like painting and panelling), which is an optional service. This type of abbreviations, can be well explained by the person who has written these abbreviations or the person in whose possession these documents were found. If such a person gives the explanation/full form of these abbreviations and the same are found convincing, looking into the facts and circumstances of the case, then the other interpretation done by the Assessing Officer to assume a higher sale price would not be justified, especially when there is no corroborative evidence to such assumptions. In this case, even the Assessing Officer did not verify his assumptions as no buyers were questioned to confirm if they paid any cash over and above the sale price mentioned in the deed. Even the Assessing Officer used internal Excel files containing budget projections to calculate the project cost of Rs. 5,000 per sq. ft. The Assessing Officer then assumed that any unit sold below this rate implies that the difference was collected in cash. However, the case of the assessee is that these Excel sheets were internal estimates and budgets, not actual sales. Moreover, neither any unaccounted cash in the possession of the assessee, nor any ‘cash book’ or informal ledger recording on-money receipts or any other evidence of the utilization of such alleged cash (e.g., unexplained investments or expenses) or any day-to-day cash trail was found or noticed during the search action. The Assessing Officer did not make any inquiries with the purchasers by issuing notices under section 133(6). Conversely, the assessee voluntarily filed affidavits from the buyers, who categorically denied paying any amount in cash over the registered value. The Assessing Officer has not referred to any single corroborative evidences containing date-wise cash receipts from any member or loose material bearing the signature of buyer confirming cash payment or other evidences which clearly bifurcates the negotiated deal into ‘cash component’ and ‘cheque component’. That even there was a timing mismatch as the units in question were booked during the years 2014-2016 at lower market rates, which were compared with a cost average calculated years later in 2021-22. Even, force is found in the contention of the assessee that the prices quoted in the loose papers/chats found from the broker would have been the resale offer/negotiation price offered by the first buyer, as these documents/chats were of a later period, even after the booking of the flats by the assessee. [Para 10.1]
  • Even the adoption of a uniform rate of Rs. 6,500 by the Commissioner (Appeals) for all units sold is purely based on assumptions and presumptions. The Commissioner (Appeals), on the one hand, has rejected the methodology and assumptions made by the Assessing Officer, on the basis of seized loose papers and held that these documents did not prove that the assessee had sold the units at the price mentioned in those documents, as those documents were dumb documents and were not corroborated with any evidence as observed above. On the other hand, the Commissioner (Appeals) himself proceeded to firstly calculate the construction cost, then applied such construction cost to all the units irrespective of their size, location and the amenities offered therein. The assessee has demonstrated that the value of a unit is determined by its floor level, direction (Vaastu), proximity to amenities, and the time of booking (early bird v. final phase). Therefore, the action of the Commissioner (Appeals) in assuming a flat rate for all units and without there being any corroborative evidence and merely on an assumption basis, cannot be held to be justified. [Para 10.2]
  • The facts in the hands clearly suggest that both the lower authorities have proceeded on the basis of their own assumptions and presumptions to assume the receipt of ‘on-money’ by the assessee on sale of units without any corroborative evidence being found during the search action or during the course of post-search inquiries. In view of the detailed discussion made herein above, the entire addition of on money made by the Assessing Officer for Rs. 20.24 lakhs for project ‘Privilon’ and Rs. 33.12 lakhs for project ‘Parijat Eclate’ is ordered to be deleted. [Para 10.6]
  • In Favour of: The assessee

[2026] 182 taxmann.com 164 (Delhi - Trib.) IN THE ITAT DELHI BENCH Spectris Technologies (P.) Ltd. v. Income-tax Officer
JANUARY 7, 2026
Section 92C of the Income-tax Act, 1961 and Section 32, read with section 37(1), of the Income-tax Act, 1961

Facts I :

  • The assessee was a domestic company and was in the business of supplying equipment and products related to material analysis. The assessee also provided installation and commissioning services, followed by after sales services, to its clients.
  • The assessee had two segments of business i.e. providing AMC Services and secondly, providing Agency & Marketing Support Service.
  • The Assessing Officer/TPO treated these two segments as part of the broad umbrella of services provided by the assessee and had treated them as one for the purposes of working out Arm’s Length Price (ALP).
  • On appeal, the Commissioner (Appeals) upheld the action of the Assessing Officer.
  • On appeal to the Tribunal:

Held:

  • The issue of TP Adjustment on the impugned services is now settled through the findings given for two previous years by the Tribunal. [Para 4]
  • Following this finding, the Assessing Officer/TPO is directed to consider the two segments of the business as presented by the assessee. Thereafter, the benchmarking has to be done with respect to the two segments of business. The TPO would analyze and consider the submissions of the assessee, including the details of comparables provided by him, and thereafter pass an appropriate order on the segmental aspects of the assessee’s business. Secondly, the TPO would also test his conclusions on the parameters and conclusions arrived at by the assessee and see if the same are within the tolerance level provided in the proviso to section 92C(2). Accordingly, with these directions the issue of working out ALP is restored to the file of the TPO/Assessing Officer. Needless to say, adequate opportunity of being heard would be given to the assessee and the comparables etc. provided by him will be fairly examined in the light of extant provisions. [Para 4.1]
  • Conclusion: Matter remanded

Facts II :

  • The assessee-company was engaged in the business of supplying equipment and products related to material analysis. It also provided installation and commissioning services, followed by after-sales services, to its clients.
  • The assessee had paid non-compete fees and claimed depreciation on the same. The Assessing Officer treated non-compete fees as capital expenditure and disallowed depreciation thereon.
  • On appeal, the Commissioner (Appeals) held that the amount paid on account of non-compete fees was capital in nature but it was also of a personal nature and hence it did not qualify for any depreciation under section 32(1)(ii). He, thus, disallowed the depreciation claimed by the assessee.
  • On appeal to the Tribunal:

Facts II :

  • The assessee-company was engaged in the business of supplying equipment and products related to material analysis. It also provided installation and commissioning services, followed by after-sales services, to its clients.
  • The assessee had paid non-compete fees and claimed depreciation on the same. The Assessing Officer treated non-compete fees as capital expenditure and disallowed depreciation thereon.
  • On appeal, the Commissioner (Appeals) held that the amount paid on account of non-compete fees was capital in nature but it was also of a personal nature and hence it did not qualify for any depreciation under section 32(1)(ii). He, thus, disallowed the depreciation claimed by the assessee.
  • On appeal to the Tribunal:

HELD – II

  • Regarding the issue of treatment being given to non-compete fees, it is found that the Commissioner (Appeals) was right at that point in time to follow the Delhi High Court in Sharp Business System v. CIT [2012] 27 taxmann.com 50/211 Taxman 576/254 CTR 233 (Delhi). However, with the latest judgment of the Supreme Court in Sharp Business System v. CIT [2025] 181 taxmann.com 657 (SC), the issue is now settled regarding the treatment to be given to non-compete fees and respectfully following the same, the Assessing Officer is directed as under: 1. The Assessing Officer must follow the said judgment in treating non-compete fees as a revenue expenditure. 2. The Assessing Officer must, with the help of the assessee, work out the treatment to be given to the said asset now in the light of the Apex Court’s judgment, with regard to the depreciation already claimed by the assessee. In this regard the portion pertaining to the depreciation issue in the Delhi High Court’s Judgment (supra) can also be utilized for giving effect to the Apex Court’s order. [Para 5]
  • Conclusion: In favour of assessee

Circulars and Notifications January 2026

Notifications

SECTION 10(23EE) OF THE INCOME-TAX ACT, 1961 - EXEMPTION - CORE SETTLEMENT GUARANTEE FUND - NOTIFIED FUND

In exercise of the powers conferred by clause (23EE) of section 10 of the Income-tax Act, 1961 (43 of 1961), the Central Government hereby notifies the Core Settlement Guarantee Fund (PAN: AAAJA3150B), set up by AMC Repo Clearing Limited, a recognized clearing corporation, with respect to specified income mentioned in Explanation (iii) of clause (23EE) of section 10 of the Income-tax Act, 1961, for the purpose of the said clause for the assessment year 2024-25 and subsequent assessment years.

  1. This notification shall be effective subject to the conditions that the Core Settlement Guarantee Fund (PAN: AAAJA3150B) set up by AMC Repo Clearing Limited shall continue to follow the conditions mentioned in clause (23EE) of Section 10, including the following-
  2. shall file a return of income in accordance with sub-section (4C) of section 139 of the Income-tax Act, 1961.
  3. AMC Repo Clearing Limited shall remain recognised as a clearing corporation by SEBI.

NOTIFICATION NO. S.O. 115(E) [NO. 06/2026/F. NO. 300197/212/2023-ITA-I], DATED 8-1-2026

Tax Calendar February 2026

07th February 2026

  • TDS/TCS Deposit: Due date for the deposit of Tax deducted/collected for the month of January 2025. However, all sums deducted/collected by an office of the government shall be paid to the credit of the Central Government on the same day where tax is paid without production of an Income tax Challan

14th February 2026

  • Due date for issue of TDS Certificate for tax deducted under section 194S/194-IA/194-IB/194M in the month of December 2025

15th February 2026

  • Form 16A: Quarterly TDS certificate (in respect of tax deducted for payments other than salary) for the quarter ending December 31, 2025

Tax News from Around the World

Scotland Announces Wealth-Related Taxes

A new “mansion tax” and private jet levies are planned, along with income tax threshold increases in the 2026-27 budget.

US State Tax Cut News

Kentucky reduces its income tax rate, expected to increase take-home pay for residents.

OECD Tax Cooperation Statements.

OECD tax cooperation statements from G7 highlight progress on international tax collaboration and global minimum taxation.