October 1, 2026, (Inside AI) — Pakistan's Federal Board of Revenue (FBR) will deploy artificial intelligence to flag tax fraud, discrepancies, and suspected evasion, according to sources familiar with the plan. A new data centre, built with World Bank support, will store and process large volumes of taxpayer records, allowing authorities to cross-check financial data and surface inconsistencies that warrant closer examination.
The initiative marks a significant escalation in Pakistan's push to modernise tax administration. The country has long struggled with low compliance and a narrow tax base. By integrating AI into compliance checks, the FBR hopes to move from manual audits to automated risk scoring. The system will draw on financial records and other taxpayer data to identify patterns that human reviewers might miss.
Why Pakistan's Tax Gap Defies Easy Fixes
Pakistan's tax-to-GDP ratio has hovered around 10% for years, among the lowest in South Asia. A large informal economy, weak enforcement, and limited data sharing between agencies have hampered collection. The FBR's new data centre aims to change that by centralising information and applying analytics at scale. The World Bank has supported similar digital tax projects in other developing countries, though results have varied.
The FBR has not disclosed the specific AI models or vendors involved. It also remains unclear how the system will handle privacy concerns or potential false positives. Tax experts warn that automated flags can burden honest taxpayers if not carefully calibrated. The FBR says the system is designed to assist, not replace, human judgement.
Read: As AI Accelerates, Governments Are Increasingly Left Behind
"AI can help identify anomalies, but the final decision must rest with trained tax officers," said a senior FBR official who spoke on condition of anonymity. "The goal is to reduce manual workload and focus resources on high-risk cases."
The move aligns with a broader global trend. Revenue agencies in India, Brazil, and Kenya have adopted AI for tax compliance, often with mixed outcomes. Success depends on data quality, legal safeguards, and staff training. Pakistan's initiative will be watched closely as a test case for AI in public finance.
The World Bank's Quiet Role In Pakistan's AI Tax Push
The World Bank's involvement extends beyond funding. The institution has provided technical assistance on data governance and analytics frameworks. A World Bank report on tax and growth notes that digital tools can improve compliance if paired with institutional reform. The FBR's project will need to address legacy systems and fragmented databases to deliver results.
Read: FTC Chair Andrew Ferguson Rejects Treating AI Agents as Independent Actors
Critics point out that technology alone cannot fix Pakistan's tax woes. Without political will to tax powerful sectors, AI may simply flag small discrepancies while large evaders remain untouched. The FBR has not announced a timeline for full deployment or how it will measure success. For now, the data centre stands as a tangible step toward a more data-driven revenue system.
