FBR Rolls Out AI-Driven Rules for Electronic Tax Scrutiny

ISLAMABAD: The Federal Board of Revenue (FBR) has introduced draft rules to deploy artificial intelligence for the electronic scrutiny of income tax returns, creating an automated mechanism to identify errors before pursuing legal or penal actions against taxpayers. The initiative marks a significant transition toward technology-led tax enforcement, designed to minimise physical contact between tax collectors and citizens while broadening the net of documented economic activities.

Formulated under draft rule ’38-B Procedure for electronic scrutiny and intimation of issues detected by the automated system’, the new framework modifies the Income Tax Rules 2002. Under this mechanism, the revenue authority’s automated system will perform cross-matching analysis by comparing figures declared in individual tax returns with vast independent databases, such as banking transactions, utility payments, and asset registries. The draft regulations have been released for stakeholder feedback and are scheduled for formal adoption following a three-day consultative window.

Upon detecting discrepancies or factual and legal inaccuracies, the system will generate electronic alerts and advance intimations directly to taxpayers through the centralized IRIS portal. Rather than initiating immediate audits or issuing punitive notices, the process provides individuals with a structured window to clarify anomalies, rectify errors, or revise returns without facing bureaucratic harassment or unexpected litigation, fundamentally reshaping procedural compliance.

The reform comes as Pakistan aggressively pursues fiscal consolidation and structural tax reforms under surveillance from multilateral lenders, aiming to enhance revenue mobilization and eliminate leakages. Following the closure of the public feedback period, the FBR plans to fully integrate the artificial intelligence modules across nationwide assessment operations, with implementation expected before the close of the current fiscal year.