AI as business partner in Pakistan’s public sector: Reimagining the PSDP

In boardrooms and policy corridors around the world, Artificial Intelligence is no longer discussed merely as a technological tool. It is increasingly recognised as a business partner — an entity that thinks alongside human decision-makers, surfaces hidden patterns in data, flags inefficiencies before they become crises, and recommends courses of action grounded in evidence rather than intuition. For Pakistan’s public sector, and particularly for the Ministry of Planning, Development and Special Initiatives, this shift in framing carries enormous practical significance. The question is no longer whether AI can help — it demonstrably can — but when and how?

The business partner paradigm

The traditional conception of software in government is that of a passive instrument: a spreadsheet that crunches numbers on command, a database that stores records, and a portal that receives applications. AI as a business partner operates on an entirely different logic. It monitors, learns, anticipates, and advises. It does not wait to be queried; it proactively identifies anomalies, models future scenarios, and calibrates recommendations as new information arrives. A finance ministry working with an AI partner, for instance, does not merely tabulate revenue shortfalls after the fact — it receives early warnings months in advance, together with scenario analyses showing the probable consequences of different corrective responses.

This distinction matters profoundly for a country like Pakistan, where the gap between policy intent and implementation outcomes has historically been wide. The challenge is rarely one of ambition; Pakistan’s planning documents are often sophisticated and well-reasoned. The challenge is execution — tracking whether funds actually flow to intended destinations, whether contractors meet timelines, whether beneficiaries are real, and whether outputs translate into the development outcomes that were promised to citizens.

The PSDP challenge

The Public Sector Development Programme, administered by the Ministry of Planning, Development and Special Initiatives, represents the federal government’s annual capital expenditure on infrastructure, social development, and economic growth. Each year, thousands of crores of rupees are allocated across hundreds of projects — roads, dams, hospitals, universities, housing schemes, and industrial zones spread across all four provinces and the federally administered territories.

Managing this portfolio is an exercise in staggering complexity. Project data is collected from multiple federal ministries, provincial governments, and executing agencies, each with its own reporting formats, timelines, and institutional cultures. Utilisation rates have historically underperformed targets, with significant portions of approved allocations lapsing or being surrendered at year-end. Cost overruns and schedule slippages are endemic. The reasons are multiple: weak feasibility studies, land acquisition delays, inter-agency coordination failures, procurement bottlenecks, and sometimes straightforward governance lapses.

An AI business partner, properly integrated into the PSDP management architecture, could begin addressing each of these failure points systematically.

Where AI can make the difference

Project Appraisal and Prioritisation. Every year, the Planning Commission evaluates hundreds of PC-I forms submitted by federal ministries for new project approvals. This process is labour-intensive and inherently subjective. Machine learning models trained on historical project data — including cost estimates, sector type, geographic location, executing agency capacity, and eventual outcomes — can score new proposals for implementation risk, flag unrealistic cost projections, and recommend sequencing priorities aligned with national development objectives. This does not replace human judgement; it sharpens it.

Real-time expenditure monitoring. AI-powered dashboards, fed directly from the Integrated Financial Management Information System and connected to the PSDP monitoring portal, provide the Ministry of Planning with a live picture of project-wise utilisation every day of the fiscal year — not just at mid-year review or at year-end. More importantly, predictive models can identify, as early as the first quarter, which projects are on track for lapse and which require immediate administrative intervention. This transforms the Ministry from a retrospective auditor into a proactive portfolio manager.

Procurement intelligence. A significant share of project delays in the PSDP originates from procurement processes — tender preparation, evaluation, and award. Natural language processing tools can review bidding documents for ambiguities, flag deviations from PPRA rules, and analyse bid patterns for indicators of collusion or artificial pricing. Benchmarking AI can compare unit costs across similar projects to detect outliers that warrant closer scrutiny before contracts are awarded.

Geographic and social targeting. Satellite imagery analytics and geospatial AI tools can verify whether physical infrastructure — a road, a school building, a water supply scheme — actually exists at the location claimed in project reports. Combined with socioeconomic data, they can also assist in identifying underserved areas that have historically been bypassed in PSDP allocations, supporting more equitable spatial distribution of development spending.

Outcome measurement. The shift from output-based to outcome-based development management is a long-standing aspiration in Pakistan’s planning framework. AI can operationalise this aspiration by aggregating data from multiple sources — household surveys, health management information systems, school enrolment records, economic activity indicators — to assess whether completed PSDP projects are actually improving lives, and by how much.

The institutional conditions for success

None of these possibilities will materialise through technology procurement alone. The Ministry of Planning has in recent years taken steps toward digital transformation — including the development of the Intelligent Project Automation System (iPAS) and efforts to link financial and physical progress reporting. These are foundations worth building upon. But three institutional conditions are essential.

First, data quality and standardisation must be treated as a strategic priority. AI partners are only as reliable as the data they process. Inconsistent project codes, duplicate entries, and manual reporting errors — all common in current PSDP data flows — must be addressed through mandatory data governance standards.

Second, human capacity must grow alongside technological capability. AI tools in the planning domain require professionals who can interpret model outputs, question their assumptions, and translate recommendations into administrative decisions. This calls for sustained investment in data literacy across the Ministry and its attached departments.

Third, political commitment to transparency must anchor the entire enterprise. AI-powered monitoring is most valuable when its outputs are acted upon — when a project flagged for poor performance actually faces consequences, and when consistent underperformers cannot shelter behind ministerial patronage. Technology without accountability is merely an expensive dashboard.

A partnership worth building

Pakistan stands at a moment when its fiscal constraints make every rupee of public investment matter more than ever. The PSDP, despite its limitations, remains the government’s most direct lever for building the physical and social infrastructure that growth requires. Transforming the Ministry of Planning’s relationship with Artificial Intelligence — from that of an occasional user of digital tools to that of a genuine institutional partner in development management — is not a futuristic aspiration. The technologies exist, regional precedents from India, Bangladesh, and the Gulf states offer lessons, and the need is urgent.

Under the leadership of Minister for Planning, Development and Special Initiatives, Professor Ahsan Iqbal, the Ministry is working aggressively towards making AI a business partner to ensure good governance, honesty about what is working, courage to act on evidence, and a genuine commitment to the citizens whose taxes fund every line in the PSDP.Great results are expected soon.

Copyright Business Recorder, 2026