This month marks six years since the “Transparent Taxation Honouring the Honest” platform was launched on 13 August 2020, and the faceless assessment and faceless appeal schemes were placed at the centre of India’s direct tax administration. The stated aims were specific and testable: eliminate the physical interface, remove territorial jurisdiction as a source of arbitrariness, introduce team-based assessment, and reduce disputes.

The question is no longer whether faceless administration was a good idea. It is now settled law. With effect from 1 April 2026, the Income-tax Act, 2025 has lifted the framework out of executive notifications and embedded it in the statute, and the Finance Act, 2026 has amended the machinery again. What could once be adjusted by a Board circular now requires Parliament. That elevation makes an honest evaluation more urgent, not less.

This letter attempts that evaluation using only figures the Government and courts have themselves placed on record. Its conclusion, stated plainly at the outset, is this: faceless administration has succeeded as a processing reform and is underperforming as a justice reform. It has made the department faster at handling paper and has not yet made it fairer, quicker or more final at deciding disputes.

I. What the record shows is working

The gains are real and should be acknowledged without qualification.

• Financial year 2025-26 was the first year in which disposals outran fresh filings. About 2.24 lakh appeals were disposed against roughly 1.79 lakh filed, cutting the standing backlog by about 34,000 appeals. That is a genuine inflection point.

• Disposal capacity has doubled. From about 1.11 lakh appeals disposed in FY 2023-24 to 2.24 lakh in FY 2025-26 is a rate of improvement no earlier reform produced.

• Downstream execution is now strong. Some 5.68 lakh of 6.29 lakh appeal-effect orders were passed in FY 2025-26, a 90.3 per cent rate. Once a case is decided, the consequential relief now actually reaches the taxpayer.

• Return processing is near-complete. Of about 8.6 crore verified returns for AY 2025-26, roughly 1.5 lakh remained pending; of 3.82 crore verified refund cases, about 1.47 lakh.

• Grievance redressal has improved materially, with 94 per cent of CPGRAMS grievances and 95 per cent of e-Nivaran grievances disposed in FY 2025-26.

Any fair assessment must start here. The machine has been built, it runs at national scale, and it works.

II. What the record shows is not working

A. The arithmetic of the backlog does not close in any reasonable horizon. Industry submissions ahead of the 2026-27 Budget placed roughly 5.4 lakh appeals, involving about Rs 18.16 lakh crore, pending before Commissioners of Income-tax (Appeals). At FY 2025-26’s net clearance of about 34,000 appeals a year, that stock takes some sixteen years to exhaust, twelve years even on the more generous reading of disposals minus filings. A taxpayer who files a first appeal today is, on the department’s own numbers, being handed a wait measured in a decade or more before the first independent look at the demand.

B. Pendency grew during the scheme, not before it. In 2017, the Parliamentary Standing Committee on Finance recorded about 2.83 lakh appeals and Rs 6.71 lakh crore locked at the CIT(A) level. By March 2022 that was 5.02 lakh appeals and Rs 14.18 lakh crore. An affidavit filed by the Revenue before the Gujarat High Court in September 2024 put pendency at about 5.81 lakh. Representations from the tax bar in early 2025 recorded 5,49,042 appeals, some pending over five years. Whatever else the faceless appeal scheme did between 2020 and 2025, it did not arrest the accumulation.

C. Assessment throughput is falling. Faceless assessments completed fell from about 2.73 lakh in FY 2023-24 to about 2.13 lakh in FY 2025-26,  a decline of roughly 22 per cent. If this reflects deliberate narrowing of scrutiny in favour of risk-based selection, that is a defensible policy and should be stated as one. If it reflects capacity constraints in the assessment units, it is a warning. At present the public record does not distinguish between the two.

D. The quality of what is being assessed remains the deepest problem. The department’s own historical record, drawn on by the Economic Survey and by OECD comparisons, shows a success rate below 30 per cent across the Tribunal, High Courts and Supreme Court — roughly two-thirds of litigated demands do not survive scrutiny. Faceless assessment was expected to improve this by replacing individual discretion with team-based review. Six years on, there is no published series showing that it has. This matters directly to taxpayers, because a stay of recovery ordinarily requires a deposit of 20 per cent of a demand that is, statistically, more likely than not to be deleted.

E. Faceless assessment is generating the writ litigation it was meant to prevent. The High Courts have, through 2025 and 2026, repeatedly set aside faceless orders on elementary grounds — replies filed and never considered, video-conference hearings requested and never granted, additions made without confronting the assessee with the material. The Bombay High Court has quashed assessments for breach of natural justice and separately cautioned against unexamined reliance on machine-generated outputs in quasi-judicial decisions. The Madras High Court has held that a statutory right to be heard cannot be defeated because a taxpayer articulated the request in a written reply rather than clicking a particular button on the portal. The Delhi and Rajasthan High Courts have quashed 2026 assessment orders on the same footing, and have had to direct the Board to make video-conferencing available in appellate proceedings at all. Every such order sends the case back to the start. The same officer-hours are spent twice, the taxpayer pays for representation twice, and the demand ages further.

F. A four-year jurisdictional failure was closed by retrospective legislation. Section 151A and the 2022 scheme framed under it required reassessment to be initiated through the faceless mechanism. Across four years, High Courts divided on whether jurisdictional assessing officers could issue notices under Sections 148A and 148 at all, and thousands of reassessments hung in the balance. The matter closed not with administrative compliance but with Section 147A, inserted by the Finance Act, 2026 and deemed effective from 1 April 2021; the Supreme Court in ITO v. Tej Partap Singh (June 2026) then set aside the High Court judgments that had quashed such notices, expressly leaving the validity and retrospectivity of the new provision open, with an interim stay and a request that the High Courts decide by 30 September 2026. The design defect was real, it persisted for four years, and the remedy was to change the law backwards rather than to fix the process forwards.

G. The anonymity that defines the scheme is under internal strain. In January and February 2026, the Principal Chief Commissioner at Chandigarh directed faceless appeal units to submit ten “quality appellate orders” each for review by a screening committee. The National Faceless Appeal Centre’s own communication of 20 January 2026 recorded that no standard operating procedure or guideline for such a review exists, and that the format remained pending with the Board. Whatever the intent, the combination is troubling: the taxpayer cannot know who decided the appeal, while officers senior to the decider can review the decision under no published rule. Anonymity was adopted to insulate adjudication from hierarchy. It cannot be allowed to do the opposite.

H. The structural gaps are known and unaddressed. There is still no binding statutory time limit for the disposal of a first appeal,  the one-year norm is advisory and, as the Punjab and Haryana High Court observed in Kulwinder Paul Singh v. CBDT, inordinate delay defeats the object of the provision. There is no time limit at all for an assessing officer to furnish a remand report, which is among the commonest causes of an appeal going quiet for years. And industry submissions place vacancies at the CIT(A) level at around 40 per cent. A scheme cannot be judged on its architecture when it is being run at three-fifths of its sanctioned strength.

III. The record in summary

Figures are as reported publicly by the Ministry, the CBDT, Parliamentary replies, court records and industry submissions, and are stated on the dates indicated. They are not always drawn on identical definitions or coverage, and are offered as a directional picture rather than a reconciled series — a limitation that is itself part of the problem.

IV. Five questions for the Finance Minister of India 

1. By which financial year does the Ministry project that pendency before the Commissioners (Appeals) and the National Faceless Appeal Centre will fall below one lakh appeals, and will that target, together with the annual net clearance required to meet it, be published and reported against? A backlog that clears in sixteen years at the current net rate is not a queue; it is a denial of remedy by arithmetic.

2. Why did faceless assessments completed decline from about 2.73 lakh in FY 2023-24 to about 2.13 lakh in FY 2025-26? Is this a deliberate narrowing of scrutiny under risk-based selection, or a capacity constraint, and what are the current sanctioned and working strengths of the assessment units, the appeal units, and the CIT(A) cadre?

3. Will the Government publish, annually and disaggregated, the number of faceless assessment and faceless appellate orders set aside by the High Courts and the Tribunal on grounds of denial of hearing or non-consideration of submissions, together with the demand involved and the cost of redoing them? Without that series, neither Parliament nor the taxpayer can tell whether the quality of orders is improving or merely being litigated at a distance.

4. Now that faceless jurisdiction is codified in the Income-tax Act, 2025, will the Government legislate a hard, non-extendable time limit for the disposal of a first appeal and for the furnishing of remand reports,  with the consequence, on breach, that recovery is automatically stayed and the 20 per cent pre-deposit released? A statutory scheme that binds the taxpayer to strict timelines and binds the department to none is not a faceless system; it is a one-sided one.

5. Under what legal authority may anonymised appellate orders be called for and screened by officers senior to the deciding authority, given the NFAC’s own record of 20 January 2026 that no procedure for such review exists — and, more broadly, does the Government accept that the faceless mandate under Section 151A went unimplemented for four years, requiring retrospective validation by Section 147A? What institutional safeguard now prevents a repeat under the 2025 Act?

V. In closing

None of the above is an argument for returning to the discretion, territorial jurisdiction and physical interface that faceless assessment displaced. That system was worse, and few who practised under it would want it back. The argument is narrower and, respectfully, more serious: a reform this significant deserves to be measured against its own stated objectives, in public, on numbers the Government itself publishes, and to be corrected where the numbers say it is falling short.

Six years is long enough to have built the machine. It is also long enough to be asked what the machine has produced. The FY 2025-26 figures suggest the corner may finally be turning. Whether it has turned far enough, and fast enough for the taxpayer whose capital and whose case have both been waiting since 2020, is a question only the Ministry can now answer.
 

Principal sources

• CBDT Chairman’s address, 167th Income Tax Day, July 2026

 appeals disposed and filed, appeal-effect orders, return and refund processing, grievance disposal, FY 2025-26.

• Written reply, Minister of State for Finance, Lok Sabha, August 2026, faceless assessments completed and appeals disposed, FY 2023-24 to FY 2025-26.

• FICCI pre-Budget submission to the Ministry of Finance, October 2025,  appeals pending and amount locked before CIT(A); vacancy levels.

• Parliamentary Standing Committee on Finance, 2017, appeals and amounts locked at each appellate stage.

• All India Federation of Tax Practitioners representation, March 2025, pendency of 5,49,042 appeals before CIT(A) / NFAC.

• Affidavit of the Revenue before the Gujarat High Court, Om Vision Infraspace (P) Ltd v. ITO, September 2024, pendency figures.

• ITO v. Tej Partap Singh (SC, June 2026); Section 147A, Finance Act, 2026; Section 151A and the E-Assessment of Income Escaping Assessment Scheme, 2022.

• High Court decisions on faceless orders and personal hearing, Bombay, Madras, Delhi and Rajasthan, 2025-26; Kulwinder Paul Singh v. CBDT (2025) 475 ITR 371 (P&H).

• Income-tax Act, 2025 (faceless jurisdiction, in force 1 April 2026); Section 144B, Income-tax Act, 1961; Faceless Appeal Scheme, 2021.

• Correspondence of the PCCIT, Chandigarh dated 16 and 30 January and 3 February 2026, and the NFAC communication dated 20 January 2026, as reported in the trade press.