AI-led bid-rigging alert system created

CCP develops system to uncover concealed patterns of cartelisation in procurement

ISLAMABAD:

The Competition Commission of Pakistan (CCP) has developed an artificial intelligence-powered system to uncover concealed patterns of cartelisation in Pakistan's public procurement market, where an estimated $60 billion is spent annually on goods, works and services.

The Bid-Rigging Detection System (BRDS) is designed to screen large volumes of procurement data and flag suspicious bidding patterns for investigation. Its development assumes particular significance as public procurement is estimated to account for approximately 15-20% of Pakistan's gross domestic product.

Historically, fragmented procurement data and the absence of automated links with corporate ownership records have constrained the systematic detection of sophisticated practices such as bid rotation, cover-bidding and participation by related companies across multiple tenders. BRDS seeks to shift detection from a largely complaint-driven and manual process towards proactive, systematic and data-led market surveillance, says a CCP statement issued on Saturday.

The initiative has also gained international recognition, with BRDS featuring in the Computational Antitrust Worldwide: Fifth Cross-Agency Report, published by Stanford Computational Antitrust. The report brings together first-hand accounts from 30 competition authorities and examines how computational tools are enabling regulators to undertake enforcement work that would otherwise be difficult – or impossible – through conventional methods.

BRDS has reached an advanced stage of development and pre-deployment preparation. Understandings have been reached with the Public Procurement Regulatory Authority, provincial procurement authorities and the Securities and Exchange Commission of Pakistan to facilitate its implementation and the institutional coordination required for effective data integration.

CCP Chairman Farid Ahmad Tarar said procurement worth trillions of rupees could not be effectively monitored solely through complaints and the manual examination of individual tenders. He said BRDS would enable the commission to identify complex and concealed bidding patterns across large datasets and direct its investigative resources towards the highest-risk cases.

He emphasised that alerts generated by the system would serve as investigative leads rather than determinations of wrongdoing. Each alert will require independent verification, evidence gathering and legal assessment under the Competition Act, 2010.

"CCP's contribution to the Stanford report also sets out practical lessons for competition authorities in developing economies, emphasising machine-readable data, inter-agency cooperation, controlled testing and legal defensibility," the statement said.