FBR’s AI-Based Tax Scrutiny: Can Tax Officers Still Initiate Assessments Without CRM Selection?

The Federal Board of Revenue is moving towards a centralized, risk-based approach to income tax assessments, with corporate and non-salary returns expected to receive greater attention under its AI-supported Compliance Risk Management (CRM) System. According to reported developments, high-risk corporate cases will be prioritized for audit selection, while salaried taxpayers are not expected to be the principal focus of the initial AI-driven scrutiny. The initiative represents a shift from conventional officer-led selection towards a system designed to identify compliance risks through taxpayer information and analytical assessment.

An important procedural development concerns the restrictions on initiating income tax assessment proceedings from September 1, 2026. According to the reported FBR directions, Inland Revenue officers cannot initiate fresh amendment or assessment proceedings, including those under sections 121, 122, 122A and 122C of the Income Tax Ordinance, 2001, unless the relevant case has been selected and assigned through the CRM System. Proceedings initiated contrary to these directions may be treated as unauthorized and potentially void from inception. Existing proceedings initiated up to August 31, 2026, may continue under the earlier framework, although fresh proceedings or notices issued thereafter are subject to the reported CRM selection requirements.

For taxpayers and tax professionals, this development introduces an additional procedural consideration when examining the validity of assessment notices. Apart from evaluating the merits of a proposed tax adjustment, taxpayers may need to verify whether the case was selected and assigned through the prescribed CRM mechanism. From a compliance perspective, businesses should also strengthen reconciliations between income tax returns, withholding statements, sales tax declarations and financial records, as inconsistencies may increase their risk exposure under data-driven scrutiny. The effectiveness of the new system will ultimately depend on transparent selection criteria, procedural safeguards and consistent implementation.