About the Centre
The Centre for Government Data Analytics (CGDA) was established in December 2025 within the Department of the Auditor-General of Pakistan (DAGP) to strengthen public sector auditing through the systematic use of data analytics and digital technologies.
As government operations become increasingly digital, audit methodologies must evolve accordingly. The Centre enables auditors to make better use of data from SAP and other government information systems to strengthen risk assessment, improve audit evidence, and support more effective oversight of public resources. It also develops analytical methodologies, provides technical support to Field Audit Offices (FAOs), and promotes the adoption of data-driven auditing across the Department.
Artificial Intelligence (AI) represents the latest stage of the Department's digital transformation. The CGDA is currently building institutional capability by enabling auditors to explore AI tools using dummy and synthetic datasets in secure learning environments. AI is not currently used in live audit engagements or operational government data. This measured approach allows the Department to evaluate AI's potential, strengthen governance arrangements, and prepare for its responsible adoption in future audit practices.
Objectives
- Transform audit methodologies from sample-based testing to population-wide analysis.
- Strengthen risk-based audit planning through data-driven insights and audit intelligence.
- Improve audit efficiency through automation and standardized analytical methodologies.
- Identify anomalies, emerging risks, and control weaknesses to support evidence-based audits.
- Build institutional capability in data analytics, AI, and emerging technologies.
- Support continuous auditing, continuous monitoring, and innovation in line with international auditing standards and best practices.
Core Functions
Audit Intelligence: Develop analytical models to identify high-risk transactions, unusual patterns, control weaknesses, and indicators of fraud, waste and inefficiency.
Enterprise Data Analytics: Acquire, integrate, validate, and analyze data from SAP Servers and other government information systems to promote continuous auditing.
Methodology: Develop analytical standards, reusable Scripts and institutional capability in data analytics.
Red Flag Identification: Provide analytical services, dashboards, risk indicators, and technical support to FAOs. Risk analyses support audit planning and execution; responsibility for audit findings, conclusions, and opinions remains with the respective audit teams.
Strategic Roadmap
- Securing Data Access of auditable entities on standard format.
- Build institutional capability in Artificial Intelligence through training and controlled experimentation.
- Evaluate AI applications to support risk assessment, anomaly detection, and audit planning.
- Develop Continuous Audit Monitoring Systems (CAMS).
- Expand predictive analytics, process mining, and network analytics.
- Strengthen data governance, cybersecurity, and responsible AI practices.
- Enhance collaboration with SAIs, INTOSAI regional organizations, academia, and development partners.
ACTIVITIES AND UPDATES
Centralized Sampling for Financial Attest Audit
FY 2025-26
Strengthening Risk Orientation, Consistency and Data-Driven Audit Planning
A Strategic Step in Modernizing Audit Planning
In continuation with the digital transformation of public sector audit, the Auditor-General of Pakistan directed CGDA to initiate a centralized sampling exercise for Financial Attest Audit for FY 2025-26 across DAGP formations. The initiative represents a practical step towards strengthening consistency, risk identification and data-driven decision-making in audit planning. Financial Attest Audit provides independent assurance on the reliability of financial statements of various governments’ entities. The exercise is conducted in compliance with Financial Audit Manual (FAM), using a consistent, standards-based methodology that strengthens accountability, transparency, and stakeholder confidence.
The exercise covered sampling for 21 Financial Attest Audits comprising of Federal, Provincial and Local Government accounts including 108 Local Government accounts in KP and 51 and 57 accounts relating to Local Government Punjab (South and North respectively), Ministry of Foreign Affairs, the Ministry of Climate Change and Environmental Coordination, Pak PWD, Geological Survey of Pakistan, etc.
The CGDA centrally conducted data extraction, cleansing, validation, population-level analytical review and sample selection using Audit Command Language (ACL), in accordance with the respective Planning Files prepared by FAOs in accordance with Financial Audit Manual (FAM).
From Sampling to Risk-Based Audit Methodology
The source data was fetched from SAP and some data was provided by FAOs. The Sampling techniques included Monetary Unit Sampling (MUS) for substantive testing, record-based sampling for Test of Internal Controls with analytical identification of High Value Items (HVIs) and Key Items (KIs). This complementary approach enables audit teams to retain statistically informed sampling while giving focused attention to transactions that are financially significant, unusual or otherwise relevant from a risk perspective.
Analytical indicators included:
- Irregular punching of Debit entries in receipt and credit entries in expenditure.
- High-value and threshold transactions
- Outliers and unusual or non-routine transactions
- Reversals and abnormal balances
- Repeated or similar transactions and unusual patterns
- Significant concentrations around vendors
- Unmatched or unreconciled head of A/cs
- Sector-specific anomalies after discussion with sector-specific auditors
- Irregular or misplaced debit or credit entries in Receipt & Expenditure, respectively
A fundamental principle of the exercise is that an analytical indicator is not, by itself, an audit finding. Analytical results identify matters requiring focused audit attention; detailed examination, applicable audit criteria and evaluation of supporting evidence remain essential before an error, irregularity, misstatement or lapses in internal control can be established.
Strategic Value for DAGP
The initiative provides a practical foundation for moving towards the whole population-based data analysis, highlighting risk areas and subsequently pulling sample from the data involving high-risk areas. By bringing analytical visibility to the whole populations before detailed fieldwork, the CGDA support audit formations in directing attention towards matters of greater financial significance or significant audit interest.
Automation of Sampling Scripts
The exercise provides a basis for developing reusable sampling scripts, standardized analytical tests and an Audit Analytics Repository. These capabilities can progressively support more consistent, transparent and data-informed audit planning across DAGP.
The strategic objective is:
“To use data and analytics not simply to analyse more transactions, but to support better audit decisions, stronger risk identification and a clearer line of sight from audit risk to audit evidence.”
The centralized sampling exercise therefore represents an important building block in DAGP’s continuing transition towards risk-oriented, evidence-focused and data-driven public-sector auditing.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key Contributions
- DAGP Data Analytics Manual: Developed step by step guide for data analytics in the DAGP, providing a standardized framework for integrating data analytics into public sector auditing.
- Budget Analytics (FY 2024-25): Identified unauthorized budget punching beyond approved Statements of Authorized Expenditure (SOAE), Excess expenditure over final grant Expenditure and releases of Rs. 65.413 billion in 13 major object heads (Employee Related Expenditure, Operating Expenditure, Civil works, Repair & Maintenance etc.). the same was conveyed to the office of CGA and thereby internal control were developed to pre-empt such cases in future.
- Payroll Analytics: Identified irregular Special Allowance 2021 payments amounting to Rs. 46.781 million during FY 2024–25 and Rs. 40.560 million in Pesh Imam stipend-related anomalies across 29 districts. The Inquiry is in process.
- Enterprise Data Analytics: Delivered enterprise-scale data extraction and analytical services to Field Audit Offices across multiple audit domains.
Contact Info
P3GW+FQQ, Constitution Ave, G-5/2 G-5, Islamabad, 44000, Pakistan
Ph.No:051-9207715
Email: cgda@agp.gov.pk