GST Reset: How Audit Pressure Forced Biggest Tax Reform
Finance Minister Nirmala Sitharaman briefs media on GST Council decisions! (Image PIB India)
From fake invoices to tighter tech controls, GST’s evolution shows how CAG-driven reforms reshaped compliance—without killing ease of doing business.
By P. SESH KUMAR
New Delhi, December 17, 2025 — The evolution of India’s GST after repeated interventions by the Comptroller and Auditor General (CAG) tells a story far more complex than a crackdown on fraud. It is the story of a digital tax system that initially trusted too much, controlled too little, and paid a heavy price in the form of fake registrations, circular trading, ineligible input tax credit (ITC) and dubious refunds. Faced with hard audit evidence and uncomfortable numbers, the Government gradually shifted course-from episodic enforcement to systemic correction.
The early years of GST were marked by a paradox. Even as collections rose and the tax base widened, the system was bleeding through invisible cracks. The architecture that was meant to be self-policing-driven by invoice matching and system-enforced discipline-had quietly morphed into a trust-heavy regime where self-declaration trumped verification. When the CAG repeatedly warned that this design choice was turning GST into fertile ground for abuse, the response was initially incremental and cautious. Over time, however, the weight of evidence became impossible to ignore. What followed was not a single dramatic reform, but a steady tightening of digital valves across the GST pipeline.
One of the most consequential shifts came with the reassertion of invoice-based control over input tax credit (ITC). By anchoring ITC eligibility to system-generated statements such as GSTR-2B and progressively restricting provisional credit, the Government moved GST back towards its original design logic. On the ground, this fundamentally altered compliance behaviour. Businesses were compelled to reconcile purchase claims with supplier declarations, turning vendor compliance into a commercial necessity rather than a moral appeal. The reform did not eliminate disputes or errors, but it decisively ended the era when ITC could be claimed as an optimistic estimate and corrected later-if at all.
Equally critical was the tightening of the registration gateway. CAG had underlined that most large-scale frauds originated not at the filing stage, but at the point of entry through the OTP based registration. Risk-based Aadhaar authentication, biometric verification in high-risk cases, geo-tagging of business premises and targeted physical verification together raised the cost of creating fake GST identities.
Field experience suggests that while genuine businesses initially faced some friction, the ecosystem rapidly adapted. The disposable GSTIN-so central to fake invoicing rackets-became far harder to create and even harder to recycle.
The expansion of e-invoicing marked another decisive pivot from trust to technology. What began as a limited reform for very large taxpayers gradually extended to mid-sized businesses, bringing a large share of B2B transactions under near real-time system oversight. This directly addressed CAG’s concern that invoices-the basic building blocks of GST-were being generated outside effective system control.
Combined with backend analytics, e-invoicing has significantly constrained large-scale circular trading and invoice layering. Fraud did not disappear, but its scale, speed and anonymity were sharply curtailed.
Parallelly, enforcement philosophy itself underwent a quiet transformation. Instead of relying predominantly on raids, searches and post-facto scrutiny, the administration increasingly turned to data-driven risk profiling.
Strengthened analytics units, network analysis of GSTIN linkages and deeper integration with income tax, customs and corporate databases began to do what the GST system was originally meant to do-spot anomalies early and act before revenue leakage multiplied.
This shift aligns closely with CAG’s insistence that prevention embedded in systems is far superior to correction through manpower-intensive enforcement.
Refunds and export-linked credits, once a favourite escape route for fraudsters, were also brought under tighter technological discipline. Automated processing combined with backend validations reduced discretion without slowing down genuine claims. Exporters experienced greater predictability, while enforcement agencies reported a decline in “claim and vanish” patterns that had plagued the early GST years.
Other leakages identified by audit-such as abuse of the composition levy and non-registration of eligible taxpayers-proved harder to fix, largely because they depend on sustained administrative vigilance rather than one-time system changes.
Data analytics is now used to flag composition dealers at risk of breaching thresholds, and third-party data is increasingly leveraged to detect unregistered economic activity. The gains here have been incremental rather than dramatic, but they reflect a maturing understanding that GST compliance must be continuously cultivated, not periodically enforced.
As these controls tightened, an inevitable concern surfaced from industry and trade bodies: was GST drifting away from its promise of ease of doing business? This tension is real and cannot be wished away.
A tax system that becomes overly rigid risks converting honest mistakes into costly disputes and compliance into a source of anxiety. Yet the experience of the past few years demonstrates the opposite danger just as starkly-convenience without control invites systemic abuse.
The emerging lesson is that the answer does not lie in choosing between facilitation and discipline, but in designing controls that are invisible to the honest and impenetrable to the dishonest. When systems auto-populate returns, restrict credits to verified invoices, flag anomalies early and allow swift, documented corrections, compliance becomes simpler, not harder.
Technology, used intelligently, reconciles ease of doing business with fiscal integrity instead of setting them against each other.
The cumulative impact of these reforms is already visible on the ground. GST fraud has not vanished, but it has become more fragmented, smaller in scale than what it could have otherwise been and harder to industrialise.
Compliance behaviour among mid-sized businesses has hardened, driven less by fear of raids and more by the inability to manipulate the system digitally. Most importantly, the direction of travel has changed. GST is no longer being treated as a conventional tax with an IT interface; it is gradually being reshaped into a genuinely digital tax system where law, technology and enforcement logic work in tandem.
Next Leg of GST Reforms
The next phase of GST reform must build on this momentum without slipping into complacency. Universal e-invoicing, tighter system enforcement of the “tax-paid-before-credit” principle, deeper real-time data integration and minimal manual overrides are essential to complete the transition. Equally important is restraint-resisting ad hoc relaxations that weaken systemic discipline for short-term relief.
CAG’s interventions have shown that audit is not an obstacle to reform but a catalyst for course correction. If policymakers continue to treat audit findings as design inputs rather than post-mortem criticism, GST can finally fulfil its original promise: a tax system that is simple for the honest, unforgiving for the fraudulent, and resilient by design. Last but not the least, let the Government publicly acknowledge CAG’s positive contributions towards strengthening the GST architecture and implementation.
(This is an opinion piece, and views expressed are those of the author only)
The Audit Watchdog Has Lost Its Bite — The CAG Must Reinvent
Follow The Raisina Hills on WhatsApp, Instagram, YouTube, Facebook, and LinkedIn
About The Author
