Pakistan’s Federal Board of Revenue is expanding the use of artificial intelligence to analyse tax returns and identify potential discrepancies in taxpayers’ declared income, assets and financial activity.
Pakistan is preparing to expand its AI-based tax monitoring system, with the Federal Board of Revenue (FBR) using data analytics and artificial intelligence to identify potential cases of under-reporting, undeclared income and inconsistencies in tax returns.
The initiative is part of the country’s broader tax digitisation and revenue-collection reforms, which aim to improve compliance while reducing reliance on manual processes.
According to an FBR official quoted by Arab News, the system will cross-check information provided in tax returns against records relating to property, bank accounts, vehicles and other financial data.
AI to Compare Tax Returns With Multiple Data Sources
Under the expanded system, AI models will analyse taxpayer declarations and compare them with information available to the tax authority.
The models will also examine tax returns against similar taxpayers and identify unusual discrepancies or patterns that may require further review.
An AI-generated flag, however, will not automatically establish tax fraud. According to the FBR official, cases identified by the system could be selected for an audit, assessment proceedings or compliance notices asking taxpayers to clarify or correct information.
The approach is designed to allow the tax authority to focus its resources on higher-risk cases rather than manually reviewing every return.
Pakistan Moves Toward Faceless Tax Administration
The AI initiative is being introduced alongside Pakistan’s wider move toward a faceless tax administration system.
The FBR established the National Faceless Centre (NFC) in Islamabad in September 2026. Under the system, tax cases are selected through a computerised, risk-based process and assigned automatically to officers, reducing direct interaction between taxpayers and individual tax officials.
The new structure separates the audit, assessment and quality-review functions among different officers. Notices, responses and hearings are also conducted electronically through the FBR’s IRIS system.
AI Already Part of FBR’s Compliance Strategy
The use of AI in tax administration is not entirely new for Pakistan.
The FBR’s Compliance Risk Management programme already uses AI and machine-learning techniques to identify risks related to non-filing, under-filing and tax evasion. The authority also uses data from multiple sources to assess compliance risks and prioritise cases for further action.
The FBR’s Lifestyle Monitoring Cell separately uses AI and social-media intelligence to compare publicly visible indicators of wealth and spending with declared income and assets.
Tax Filing Deadline Extended to October 15
The expansion of AI-based checks comes as Pakistan’s tax authorities have also extended the deadline for filing Tax Year 2026 income tax returns.
The FBR extended the deadline from September 30 to October 15, 2026, following requests from business organisations and tax practitioners.
The additional time gives taxpayers more opportunity to complete their filings while the FBR prepares to apply its expanded data-analysis systems to tax returns.
Technology at the Centre of Tax Reform
Pakistan has increasingly turned to automation, artificial intelligence and data integration as part of its efforts to improve tax collection and broaden the tax base.
In June, the National Assembly’s Standing Committee on Finance and Revenue was briefed on plans involving a Central Data Hub supported by AI, algorithm-based dispute resolution, digital invoicing and greater use of third-party data for risk assessment and compliance monitoring.
The government has also said that the first phase of its AI- and technology-based faceless Inland Revenue system would become operational from October 1, with a complete rollout targeted for 2027.
What the AI Tax System Means for Taxpayers
The expanded use of AI could allow the FBR to process significantly larger volumes of tax information and identify potential inconsistencies more quickly.
At the same time, the technology will not replace the legal assessment process. A discrepancy identified by an algorithm is a risk indicator rather than proof of tax evasion, and cases can proceed to audit or assessment where further examination is required.
As Pakistan continues to digitise tax administration, AI and automated data analysis are expected to play a growing role in how the FBR identifies compliance risks and manages tax cases.
