57th GST Council’s Arrest and Prosecution Reforms: Do They Apply to Pending Cases?

Summary: The 57th GST Council Meeting held on 8 October 2026 recommended significant changes to the enforcement framework under the Central Goods and Services Tax Act, 2017. The most consequential proposal is the complete withdrawal of the power of arrest under Section 69, accompanied by an increase in the prosecution threshold from ₹1 crore to ₹5 crore, narrowing of specified offences under Section 132 and greater judicial discretion in sentencing. The Council also recommended reducing general penalties, restricting interception of goods in transit and introducing procedural safeguards against blocking of input tax credit. Together, these proposals indicate a shift towards less coercive GST administration while retaining enforcement against serious tax evasion. The central legal question concerns the application of these beneficial changes to pending investigations, prosecutions and appeals. While withdrawal of a statutory arrest power would generally affect the exercise of that power after commencement, its consequences for persons already arrested or subject to existing judicial orders require separate consideration. The proposed reductions in punishment may support claims for the benefit of a lighter sentencing framework, particularly under the principles recognised in T. Barai v. Henry Ah Hoe. However, the effect of increased prosecution thresholds and omitted offences on pending proceedings raises more complex questions under Section 6 of the General Clauses Act, 1897, and the principles discussed in Fibre Boards (P) Ltd. v. CIT. The article examines these issues through the constitutional protection of personal liberty, the rule of law and the distinction between substantive criminal liability and procedural powers. Its central argument is that the transitional provisions of the proposed legislation will be crucial in determining whether the reforms benefit persons already facing GST proceedings. The Council’s recommendations do not themselves amend the law, and the final legal position will depend upon the enacted provisions, commencement dates and judicial interpretation.

A Softer GST, but for Whom? The Reach of the 57th Council’s Arrest and Prosecution Reforms to Pending Cases

“The power to tax involves the power to destroy.” – Chief Justice John Marshall, McCulloch v. Maryland (1819)

For nine years, a summons under Section 70 of the CGST Act has carried an unwritten postscript. A taxpayer who walked into the GST Office to give a statement could not always be sure they would walk out. Section 69, inherited from the excise and service tax regimes, placed the power of arrest in the hands of the very officers investigating the tax. The line between a GST Officer and a Police Officer was getting thinner and thinner. For a genuine business, arrest carried precisely the destructive power Chief Justice Marshall warned of: a businessman in custody cannot run a factory, meet a payroll, or reassure a banker, whatever the case’s eventual outcome. The courts have spent much of the intervening period defining its limits, from the Telangana High Court in P.V. Ramana Reddy to the Supreme Court in Radhika Agarwal, and Sunil Biyani.

On 8 October 2026, the GST Council recommended that the power be abolished altogether. It did not stop there. It proposed to raise the bar for prosecution fivefold, to prune the offences that may be prosecuted, and to let courts punish by fine alone.

It is tempting to read this as the end of fear in GST administration. The more accurate reading is narrower. A law may be softened for the future and remain unchanged for the past. The real question is not whether the State has relented, but whether its relenting will reach the client who is already summoned, already arrested, or already on trial.

That GST enforcement is softening is not in doubt. It is softening perceptibly and deliberately, but in its coercive and penal edge rather than in its appetite for revenue. If the 56th Council in 2025 was about rate rationalisation, the 57th is about the temperament of administration. Read as a whole, the recommendations move the system away from enforcement by custody, penalty and interception, and toward enforcement by data, matching and automation.

The Council’s own choice of heading is telling. In the press release, the withdrawal of arrest, the rationalisation of prosecution and the e-way bill reforms do not appear under enforcement, investigation or anti-evasion. They appear under the heading “Ease of living and doing business”. The words “ease of living” deserve emphasis. They recognise that arrest under a fiscal statute was never only a cost to commerce; it fell on individuals, their families and their liberty.

The signals of softening are consistent across every stage of a dispute:

The counter-signals are few, but one deserves attention: the proposed validation clause, which would revive notices that courts have set aside for covering multiple financial years, and do so retrospectively. The Council has thus shown that it is prepared to reach back into pending matters where the Revenue’s interest requires it. With respect, the same willingness would be well placed on the taxpayer’s side. If retrospection can restore a defective notice, it can equally relieve a person facing prosecution for conduct that the State itself no longer regards as criminal. On balance, the recommendations withdraw coercive powers while strengthening system-based detection. Taxpayers facing technical or procedural defaults stand to gain the most.

The softening, then, is real. But a reform announced for the future does not, by that fact, reach matters already in the pipeline. That is the central question of this article: does the softening reach the person already under investigation or already on trial? The answer, it is submitted, differs according to the limb of reform engaged, and on the limbs that matter most it will be settled by the transitional clause of the amending Bill.

The question and the answer in brief

At its 8 October 2026 meeting, the 57th GST Council recommended completely withdrawing the power of arrest by omitting Section 69 of the CGST Act, 2017, and substantially recasting Section 132. The prosecution threshold is to rise from ₹1 crore to ₹5 crore, three offence-creating clauses are to be pruned, and sentencing is to become discretionary.

These are recommendations, not law. The Council’s note to the press release states that its decisions take effect only through the relevant notifications, circulars and amendments to the statute. Sections 69 and 132 therefore remain in their present form until Parliament (and each State legislature, for the SGST Acts) amends them. The Government has not yet announced when the amending Bill will be introduced; it may be introduced in the Winter Session of Parliament.

Against that background, the reach of the reforms to pending matters may be stated in three tiers, each developed below:

• Withdrawal of arrest reaches every matter, pending or future, from the date the amendment commences, because a procedural power once withdrawn cannot be exercised for past conduct.
• Reduced punishment reaches every accused not yet sentenced, by force of the beneficial-amendment principle in criminal law.
• The higher threshold and the narrowed offences reach pending prosecutions only if the amending Bill so provides, or if courts are persuaded that Section 6 of the General Clauses Act must yield.

The third tier is where the softening is most needed and least assured. It is therefore where the transitional clause, and the representations made before it is drafted, will matter most.

What is proposed for Sections 69 and 132

The arrest power goes entirely; the offence provisions are narrowed and softened but survive. The table sets the present provision against the proposal, drawn from the PIB release of 8 October 2026 and the Ministry’s FAQs of 9 October 2026.

Two features deserve emphasis. First, the core offence of fraudulent ITC survives in the redrafted clause (c): the reform narrows the periphery of criminal liability, not its centre. Second, the shift from “imprisonment and fine” to “imprisonment or fine or both” converts every conviction into a matter of judicial discretion. Each of these changes reaches the pending case by a different route, which the following sections trace.

Tier one: arrest and the ongoing investigation

Until the amendment is notified, Section 69 remains good law. An arrest made today is not rendered illegal by the Council’s decision. It must, however, satisfy the safeguards the Supreme Court articulated in Radhika Agarwal v. Union of India (2025): “reasons to believe” must rest on material, not suspicion; the grounds must be furnished; and the protections of the BNSS, including anticipatory bail, apply. CBIC’s own arrest guidelines (Instruction No. 02/2022-23 [GST-Investigation], dated 17.08.2022) continue to bind officers.

That said, the Council’s unanimous policy judgment that arrest is unnecessary under GST is a weighty consideration. With respect, it would sit uneasily with that judgment for the Department to authorise fresh arrests in the interregnum, save in the gravest cases. Counsel may properly place the recommendation before a court hearing an anticipatory or regular bail application, not as binding law but as the clearest statement of the State’s own view on the necessity of custody.

After the amendment, no fresh arrest can be made for a GST offence, whatever the date of the alleged conduct. Arrest is a procedural power, not a right accrued to the Revenue; once the power is withdrawn, it cannot be exercised for past conduct any more than for future conduct.

The investigation itself continues. Summons under Section 70, search and seizure under Section 67, and the adjudication of tax, interest and penalty are untouched. The omission removes the coercive edge of an investigation, not its existence.

Persons already arrested or on bail are not discharged automatically. Their liberty will thereafter be governed by the court seized of the complaint. Bail conditions that presuppose a continuing power to re-arrest will merit modification.

One caution remains. Section 132(5), which makes offences above ₹5 crore under clauses (a)–(d) cognizable and non-bailable, must be amended in step; otherwise an argument may be advanced that police arrest survives through the BNSS.

On this tier, therefore, the softening reaches every pending matter in full, though only from the date of commencement and only as regards arrest under the GST law.

Tiers two and three: pending prosecutions

In prosecutions, the reach of the softening is uneven. Reduced punishment travels to the pending case almost as of right; the higher threshold and the narrowed offences do not.

1. Reduced punishment: the accused should get the benefit. Article 20(1) forbids a heavier penalty retrospectively; it does not forbid a lighter one. In T. Barai v. Henry Ah Hoe (1983) the Supreme Court held that where an amendment reduces the punishment for an offence, an accused not yet sentenced is entitled to its benefit. The softened slabs, the option of a fine alone, and the removal of the six-month minimum under Section 132(3) should therefore be available in every trial in which sentence has not been passed when the amendment commences. Where a conviction is in appeal, the same principle supports a plea for modification of sentence.

2. The ₹5 crore threshold: the contested ground. A prosecution pending for an amount between ₹1 crore (or ₹2 crore) and ₹5 crore will be the battleground. The Revenue may invoke Section 6 of the General Clauses Act, which preserves pending proceedings when a provision is repealed, and Fibre Boards (P) Ltd. v. CIT (2015) treats an “omission” as a repeal for this purpose. The accused will answer that conduct the legislature no longer considers criminal should not continue to be prosecuted, and that Section 6 yields where a “different intention appears”.

The drafting of the Bill will likely decide this. It is instructive that the Council expressly directed that the new ₹10,000 threshold for show cause notices apply to pending notices and appeals, yet said nothing comparable about the prosecution threshold. If the Bill mirrors that silence, courts may infer that the legislature deliberately left pending prosecutions alone.

3. Omitted and narrowed offences. Prosecutions founded on clause (i), on the “evades tax” limb of clause (e), or on the “in any other manner deals with” limb of clause (h) face the same Section 6 question, but the accused’s position is stronger: the Ministry’s FAQs record that clause (i) is being omitted because it is difficult to quantify and open to subjective interpretation. A Department that has itself characterised an offence in those terms may, respectfully, find it difficult to justify pressing a prosecution under it. A prayer for withdrawal under Section 321 of the Criminal Procedure Code (now Section 360 BNSS), or discharge, merits serious consideration.

For fresh sanction proposals in the interregnum, CBIC’s prosecution guidelines (Instruction No. 04/2022-23 [GST-Investigation], dated 01.09.2022) already counsel restraint and ordinarily require adjudication to precede prosecution. A sanctioning authority that launches a below-₹5-crore prosecution in the weeks before the amendment would be acting within its powers, but contrary to the evident direction of policy.

The rule of law dimension

In Dicey’s classic formulation, the first meaning of the rule of law is that no one may be punished, or made to suffer in body or goods, except for a distinct breach of law established in the ordinary manner before the ordinary courts. Section 69 sat uneasily with that ideal. It allowed liberty to be curtailed at the instance of the investigating officer, often before the tax itself had been adjudicated and long before any court had found an offence proved. The omission of Section 69 returns the question of liberty to the courts, where Article 21, read with Maneka Gandhi v. Union of India (1978), places it: under a procedure that is fair, just and reasonable, and decided by a judge rather than by a party to the dispute. The move from arrest by the Department to prosecution on complaint before a Magistrate is the rule of law at work.

The rule of law also asks for equality and certainty. If two persons commit identical conduct below the new threshold, and one is prosecuted only because the complaint was filed before the amendment commenced, the difference between them rests on timing rather than culpability. Article 14 does not forbid a cut-off date, and courts have repeatedly upheld them. But a legislature committed to trust-based taxation may, with respect, prefer not to preserve a distinction that the State itself no longer considers justified. Here, too, the transitional clause is where the principle will be honoured or set aside.

Conclusion

The 57th Council has softened GST enforcement, and the omission of Section 69 is its clearest expression. In respect of arrest at least, Marshall’s warning has been heeded: the power to tax will no longer carry with it the power to detain. But softening announced for the future is not the same as relief for the case already pending. On arrest, relief will be complete from commencement. On sentence, it should follow as of right for any accused not yet sentenced. On the threshold and the narrowed offences, which are where most pending prosecutions sit, it depends on a single provision not yet drafted: the transitional clause of the amending Bill. Whether the reform is merely prospective or genuinely remedial will be decided there, and it is there that the attention of taxpayers should now turn.

Sources

• PIB: Recommendations of the 57th Meeting of the GST Council, 8 October 2026
• FAQs on rationalisation and decriminalisation of arrest and prosecution provisions, 9 October 2026 (via TaxGuru)

This article is based on the Council’s recommendations and the Ministry’s FAQs; the position should be re-examined against the amending Bill once introduced.

******

Disclaimer: The article is only for information purposes and is not a legal opinion of any sort. The views, thoughts, and opinions expressed in the text belong solely to the author, and these are purely for educational purposes, not a substitute for actual legal provisions. For any query, the author can be reached at [email protected].