Pakistan’s Federal Board of Revenue (FBR) is expanding the use of artificial intelligence based checks on tax returns to identify undeclared income and assets as part of efforts to strengthen tax compliance and improve revenue collection. The initiative is being introduced as part of Pakistan’s broader economic reform commitments under an International Monetary Fund (IMF) backed program, with the tax authority focusing on technology driven methods to improve monitoring and enforcement. Pakistan has long faced challenges in expanding its tax base, with a significant portion of economic activity remaining outside the formal tax system. Through the adoption of AI based analysis, FBR aims to improve the accuracy of tax assessments and create a more data driven approach to revenue administration.
The FBR plans to use artificial intelligence models to cross-check taxpayer declarations against information available from multiple sources, including property records, banking information, vehicle ownership data, and other government databases. A pilot program involving a limited number of income tax and sales tax returns has already been completed, and the authority intends to expand the use of these models for income tax filings for Tax Year 2026 and future sales tax returns. The AI systems will review submitted returns against tax law requirements, compare declarations with information from similar taxpayer categories, and identify possible inconsistencies between reported details and external records. Officials have stated that an AI flagged discrepancy will not automatically confirm any violation but will indicate cases that may require additional review. Depending on the findings, flagged returns could be selected for audit, assessment procedures, or compliance notifications requesting taxpayers to update their information where required.
The introduction of AI based tax checks is part of FBR’s wider efforts to modernize tax administration through digital technology and automated data analysis. Faisal Sattar, chief executive of Pakistan Revenue Automation Ltd., the public sector technology company supporting FBR, said the authority has established the required infrastructure to collect and analyze information from various sources. This includes tax return records along with property data, vehicle information, travel related records, and banking details. According to Sattar, the volume of available information makes traditional manual scrutiny difficult, while automated systems can help identify patterns and potential differences more efficiently. The AI models are being developed to compare information across different databases and provide insights into whether declared details align with available records. The approach is intended to improve consistency in identifying cases that require further examination while supporting a more technology enabled tax management system.
The AI based monitoring initiative is being introduced alongside additional reforms in FBR’s audit process. On September 25, FBR announced the establishment of a National Faceless Center in Islamabad, where a computerized risk based system will be used to select audit cases and assign them automatically to officers. The updated framework is designed to reduce direct involvement between taxpayers and individual officials while promoting a more standardized audit process. Separate officers will be responsible for conducting audits, determining assessments, and reviewing outcomes before final decisions are issued. FBR aims to gradually expand automation beyond identifying discrepancies by developing systems that can also assist taxpayers in resolving filing issues. Faisal Sattar said the longer term objective is for AI systems to provide possible settlement guidance when differences are identified, helping taxpayers understand how issues can be corrected. The technology focused approach supports Pakistan’s goal of improving tax administration as the country works toward achieving its federal tax collection target of Rs15.26 trillion for the financial year ending June 2027.
Source Intelligence Layer: 1 | 2
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
