GST Reforms: Wider Tax Credits, Faster Refunds and Fewer Arrest Powers—What Changes?

By P. SESH KUMAR

The proposed GST changes seek to ease compliance, expand refund eligibility and rationalise enforcement, but their impact will depend on statutory amendments, clear rules and safeguards for genuine taxpayers.

New Delhi, October 12, 2026 — GST’s promise is that tax paid on business inputs should ordinarily flow through the chain rather than become embedded in costs. The Council proposes removing specified restrictions on outdoor catering, health and life insurance, telecommunications towers, pipelines outside factory premises, free samples and certain goods destroyed or written off on expiry as required by law. These measures address places where the tax’s neutrality has been interrupted.

But the language “credit will now be available” is premature. Eligibility will depend on the enacted amendments and their conditions. Nor should a proposal concerning destruction or write-off linked to expiry and legal requirements be expanded into an unrestricted entitlement for every stock loss.

The proposed limited same-line-of-business credit for restaurant and catering services, hotel accommodation up to Rs 7,500 per unit per day, and gym or fitness services is a separate reform. It would extend a treatment already available in specified transport-related services. It is not unrestricted credit on every expense incurred by those sectors, and passenger transportation should not be described as an entirely new concession in this package.

The economic rationale is persuasive: where a service is acquired and supplied onward, complete denial of credit can introduce an additional embedded tax burden. The drafting challenge is to define the permitted chain precisely enough to prevent artificial reclassification while accommodating genuine agency and resale arrangements.

The refund expansion is more ambitious. For inverted-duty claims, accumulated credit attributable to input services is proposed to qualify for credit availed from November 1, 2026. Capital-goods credit would enter eligible zero-rated and inverted-duty refunds for credit availed from April 1, 2027, with refund spread over sixty months.

Here a crucial distinction is needed. Exclusion from an accumulated-credit refund does not necessarily mean exclusion from ordinary credit availment or utilisation. The reform concerns access to cash refund where eligible credit accumulates; it should not be described as correcting a universal denial of machinery credit.

Spreading capital-goods refunds over sixty months improves access to recovery but leaves a substantial financing burden during the intervening period. Calling that period the asset’s “working life” is also too categorical: it is a prescribed allocation period, not an empirical finding about every machine. Rules must address disposal, changes in use, exempt activity, business closure and interaction with ordinary credit utilisation so that the same tax is neither stranded nor recovered twice.

The genuine buyer remains the unfinished business

There is a possibility that a Committee of Officers will examine protection for genuine purchasers and report within three months. That proposal is not recorded in the press release and should be identified as a briefing-note statement, pending verification in an official record.

The underlying issue is central. A purchaser may hold an invoice, receive goods and pay the supplier, yet face a credit dispute because of the supplier’s conduct. CAG’s 2025 findings include credit passed on without the supplier remitting tax, demonstrating the reality of this control problem. They do not, however, establish that every affected purchaser acted dishonestly-or that purchaser protection should be abandoned.

A credible reform should define reasonable purchaser diligence, identify the evidence that rebuts an allegation of collusion, and explain when recovery should be pursued against the defaulting supplier rather than the recipient. Otherwise, wider credit eligibility can coexist with continuing uncertainty over whether credit already taken will survive.

This is where the rhetoric of trust meets its hardest test: whether the system can distinguish participation in fraud from exposure to someone else’s default.

Enforcement: less coercion, not less accountability

The recommendation to omit section 69 and withdraw GST arrest powers is a major recalibration. The Council also proposes raising the prosecution threshold from Rs 1 crore to Rs 5 crore, narrowing specified offences and rationalising punishment. The official release does not spell out every sentencing assertion such as its categorical account of minimum punishment and judicial discretion. Those details must await the proposed statutory text.

The policy justification is that tax recovery and criminal punishment serve different purposes. Routine errors, classification disputes and delayed compliance should not automatically acquire the character of criminal misconduct. But better detection does not make deterrence obsolete. Detection contributes to deterrence; it does not replace it.

Nor does withdrawal of GST arrest powers mean the disappearance of prosecution or immunity under other applicable laws. The package expressly seeks to retain an effective response to fraud. Its success will depend on preserving evidence, tracing controlling persons and pursuing well-founded cases while preventing coercion from becoming a substitute for adjudication.

The Rs 10,000 minimum threshold for show-cause notices under the specified demand provisions is similarly welcome. It concerns the combined tax amount, including the components identified in the release, and proposed treatment of pending below-threshold matters. It is not a blanket exemption from every obligation or enforcement mechanism under GST. The reduction of the maximum general penalty under section 125 from Rs 25,000 to Rs 10,000 must also remain distinct from specific penalties elsewhere in the legislation.

Disputes: better hearings-or better-worded bad orders?

The proposed common guidance on notices, adjudication and appeals addresses a familiar danger: mechanically generated allegations followed by formulaic decisions. The release specifically calls for proper invocation of fraud grounds, adherence to natural justice and personal hearings. It also proposes lower penalties in specified non-fraud cases, removal of the minimum penalty floor in those cases, and a Rs 40 crore combined cap on pre-deposits in penalty-only appeals.

Common standards will help only if they improve reasoning, not merely formatting. A good notice must identify the transaction, evidence, legal basis and computation. A good order must engage with the defence. Standardisation should prevent arbitrariness, not industrialise it.

The rule 86A proposal offers a related safeguard: an objection and personal hearing before the officer decides the objection to credit blocking. That is not necessarily a universal pre-blocking hearing. Its value will depend on prompt consideration, reasoned decisions and correction when the basis for blocking disappears.

One recommendation deserves particular scrutiny: statutory validation of notices invalidated because they covered multiple financial years. Administrative convenience is not, by itself, a sufficient answer to the legal questions this raises. The amendment must identify the defect being cured and preserve meaningful protection concerning limitation, intelligibility of allegations and opportunity to respond. A programme presented as reducing litigation should be careful not to reward defective procedure at the taxpayer’s expense.

Goods in transit: a common market needs a common enforcement discipline

The Council proposes intelligence-based interception authorised by an officer not below Joint Commissioner. Inspection and further detention or seizure would ordinarily be tied to a State where the supplier or recipient is located or registered, with no interception in transit States. Confiscation under section 130 would not apply to goods or conveyances in transit.

The origin-and-destination formulation is an accessible shorthand, but it is not identical to the release’s supplier-or-recipient location or registration test. It also omits a significant exception: goods without an e-way bill, or without documentation showing origin or destination, may be inspected, detained or seized irrespective of jurisdiction.

The objective is to prevent repeated roadside intervention from undermining the common market. The challenge is making intelligence travel faster than the truck. Authorities will need interoperable alerts, a record of prior inspections and clarity about which jurisdiction must act. Advance authorisation should impose discipline without becoming an excuse for delay in responding to credible information.

The missing-document exception must remain usable against genuine evasion without turning every technical defect into a route around the protection against transit-State interception.

Small businesses, digital commerce and service exports

The proposed Annual Return Quarterly Payment scheme recognises that frequent returns can impose disproportionate costs on smaller businesses selling exclusively to unregistered persons. Its turnover ceiling is Rs 5 crore in the preceding financial year, and approval remains in principle.

Its eventual design must address an occasional B2B transaction, turnover crossing the ceiling, entry and exit, records required during the year and treatment of input credit. An annual return does not necessarily mean annual visibility for the administration, nor should it mean postponed discovery of a year’s accumulated mistakes. The separate proposed late-fee waiver is narrower: eligible taxpayers must file the delayed return by the end of its due month.

For small e-commerce goods suppliers, proposed rule 14B would permit use of an operator’s warehouse as the principal place of business in a State where the seller lacks physical presence, subject to conditions and the specified credit-passing ceiling. This addresses a real expansion barrier. But a seller does not ordinarily need registration in every State merely because customers there purchase its goods. The reform concerns circumstances requiring a registration associated with operations or stock in another State, not a universal destination-registration obligation.

Implementation must resolve warehouse consent, stock accountability, multi-platform selling and transition when eligibility is exceeded. Simplification should not merely shift the gatekeeping power from the tax officer to a dominant platform.

The service-export recommendations would remove the distinct-establishment restriction in the export definition and shift specified services performed on goods physically made available by the recipient toward the default recipient-location rule. These could improve access to export benefits for branch-related service arrangements and work on foreign clients’ goods. But export status would still depend on the remaining statutory conditions. It would be inaccurate to promise automatic export treatment for every research, processing or contract-manufacturing arrangement.

The proposed SEZ/FTWZ deeming provision similarly seeks certainty for supplies to overseas buyers delivered within those zones, subject to the stated payment conditions. Each reform should be accompanied by examples showing where the benefit applies–and where it does not.

GST Reform: Rationalisation or Admission of Failure?

“No rate changes” is too sweeping

The assertion in some quarters that no rates changed requires qualification. The meeting did not announce another general restructuring of the principal rate framework. Nevertheless, the release expressly contains proposed rate and exemption measures: a 5 per cent option for specified electric-vehicle passenger transport and rental, 5 per cent treatment for specified delivery services, nil treatment for psyllium seeds, alignment for retreaded tractor tyres and several service exemptions.

It also proposes reverse charge for specified waste and scrap supplied by unregistered persons, and 2 per cent TDS on specified registered-to-registered supplies. Other recommendations concern classification, used-vehicle margin-scheme credit, R&D self-certification, highway concessions, shipping establishments, intellectual property and banking fund-transfer pricing. These are not all mere clarifications; some change liability, eligibility or compliance obligations.

The accurate description is therefore: no broad rate reset, but targeted rate, exemption, classification and collection-mechanism changes. An annual rate-review calendar is not stated in the press release.

What CAG has highlighted–and what it has not

The connection to CAG is substantial but must not be overstated. Its 2019 report identified incomplete invoice matching, registration-validation defects, payment reconciliation weaknesses and failures to translate legal requirements accurately into software. Its 2022 report addressed return-data inconsistencies, delayed and erroneous refunds, deficient pre-sanction risk controls and inadequate post-audit. Its 2025 report highlighted substantive compliance deviations and weaknesses in departmental monitoring, including cancellation, suspension, notices and recovery. These findings closely correspond to several operational problems the Council now proposes to address.

However, the reviewed reports do not establish that CAG advocated every element of this package. They should not be cited as an endorsement of arrest abolition, the precise prosecution threshold, every blocked-credit liberalisation or the proposed export-definition amendments. Those require their own policy and legal justification.

Historical audit observations also cannot be treated as a current census of defects. Selected samples do not establish national incidence; amounts identified in audit are not interchangeable with final recoveries or adjudicated losses. Nor does correspondence between an audit finding and a reform prove that the audit caused the decision.

CAG’s enduring contribution is a more demanding proposition: a digital tax system must be auditable as a system. Clean dashboards do not compensate for dirty data, and automated decisions do not become lawful merely because no officer clicked the button.

Lessons and the way forward

Implementation should begin with a publicly accessible statement of

• what requires legislation,
• what requires rules or notifications,
• what awaits consultation and
• what remains conceptual.

Effective dates, transitional treatment and dependencies should be stated together. This would prevent announcement-driven compliance errors and make progress measurable.

The return redesign should be tested across difficult cases before mandatory rollout: credit notes, earlier-period corrections, reverse charge, partial receipts, registration cancellation and counterparties responding at different times. CAG’s earlier findings concerning deficient requirement specifications, inadequate validation and weak change management make this a substantive necessity, not an IT formality.

Risk-based administration must also be contestable. Sensitive detection methods need not be published in full, but a taxpayer affected by an adverse classification should receive an intelligible explanation and a route to correct inaccurate information. Officers need training and incentives that reward legally sustainable outcomes, not notice volumes. The relevant measures are fewer unnecessary demands, quicker actual refunds, timely restoration of registrations, lower compliance costs and better action against genuine evasion.

Independent review should track both facilitation and revenue protection. Faster refunds should be evaluated alongside duplicate-payment prevention and post-audit quality. Automatic registration should be examined alongside identity integrity. Reduced roadside intervention should be assessed alongside successful intelligence coordination. Genuine-buyer protection should advance alongside recovery from deliberate supplier default.

The 57th meeting’s promise is not that GST will become effortless. It is that effort will be directed where it belongs: toward accurate reporting and genuine compliance, rather than navigating avoidable procedural friction. The package will succeed if a lawful transaction moves smoothly, an honest mistake can be corrected, a refund arrives when due and a fraud case is pursued on evidence rather than pressure.

We do not merely need GST to collect more efficiently. It needs GST to behave more fairly. The Council has proposed a significant move in that direction. Whether taxpayers experience it will depend on what happens after the press conference–in the statute book, in the software and at the officer’s desk.

GST Council’s Big Shift: Faster Refunds, Easier Compliance and Fewer Enforcement Powers

(This is second of the two-part series. Views expressed are the author’s own.)

Follow The Raisina Hills on WhatsApp, Instagram, YouTube, Facebook, and LinkedIn

About The Author