New Delhi | Businesses claiming Input Tax Credit (ITC) refunds under the Goods and Services Tax (GST) regime are likely to face a far more stringent and technology-driven process in the coming months. With the GST portal introducing a revised refund filing mechanism, companies will now be required to undergo detailed digital verification for every invoice submitted as part of a refund claim. Experts believe the move could significantly curb fraudulent refund claims, but it may also increase compliance pressure and preparation timelines for businesses.
The new framework is expected to particularly impact exporters, companies operating under the inverted duty structure, and businesses that regularly claim refunds on unutilised ITC. Under the revised process, taxpayers must now prepare Annexure-B through an offline utility and upload it in JSON format along with their refund applications.
FCRF Academy Launches Premier Anti-Money Laundering Certification Program
A JSON file is a structured digital format that allows the GST system to automatically read and validate invoice-level information. Earlier, most refund applications were examined manually after submission. Under the updated mechanism, however, automated scrutiny will begin at the filing stage itself.
The revised filing system requires businesses to furnish far more granular information for every invoice included in the claim. This includes the nature of inward supplies, document type, total ITC claimed, blocked credit details, eligible and ineligible ITC amounts, and the relevant GSTR-2B period associated with each invoice.
Tax professionals say that until now, mismatches between refund claims and GSTR-2B records were usually identified later during departmental review, often leading to notices, clarification requests and prolonged delays in refund processing. With the latest changes, the GST portal itself may immediately flag discrepancies during submission, making the process more rigorous and data-sensitive.
As a result, businesses will now need to carry out deeper reconciliation of purchase registers, GSTR-2B records, electronic credit ledger balances and refund calculations before filing claims. Experts warn that even minor mistakes — such as incorrect tax periods, mismatched invoice entries or claims involving blocked credits — could lead to automated system flags and possible delays in refund approvals.
The impact is expected to be more significant for companies with large procurement networks and multiple vendors, where reconciliation gaps are more common. Industry bodies have expressed concerns that the transition phase could initially slow down refund processing because of technical mismatches and enhanced validation checks.
According to tax experts, the government’s broader objective is to strengthen audit trails and crack down on fraudulent ITC refund claims. Over the past few years, authorities have uncovered several large-scale GST fraud cases involving fake invoices and shell companies used to siphon off crores of rupees through fraudulent refunds. The new system is being viewed as part of a wider shift towards data-driven and automated tax administration.
Experts also believe the updated process will push businesses to adopt stronger accounting systems and more transparent compliance practices. Companies may now need to monitor vendor GST filings more closely, since discrepancies at the vendor level could directly affect refund eligibility and processing timelines.
Industry stakeholders say that while the system could eventually speed up genuine refund approvals through automated verification, businesses will need to invest significantly in internal controls, reconciliation mechanisms and compliance infrastructure during the transition period.
The move is being seen as a major step towards a more technology-oriented and surveillance-driven GST ecosystem. Going forward, businesses may have to focus not only on timely tax payments but also on maintaining highly accurate, real-time financial and compliance data to avoid disruptions in refund processing.
