Punjab Assembly begins evaluation of AI use in public services

Lahore witnessed the commencement of parliamentary examination into artificial intelligence and digital governance by the Punjab Assembly's Special Committee No. 13 on Thursday. The committee, chaired by Speaker Malik Muhammad Ahmad Khan, is the first parliamentary committee in Pakistan and the South Asian region specifically focused on these subjects.

The inception of the committee came after a three-day workshop in Islamabad, organized by UNDP Pakistan and the UK's international development organization, from September 2 to 4.

The workshop delved into the potential benefits and risks of artificial intelligence and digital technologies, as well as the principles of responsible governance. The committee is set to initiate an inquiry program concerning the use of AI in Punjab's public services, with particular emphasis on transparency, accountability, citizens' rights, data security, remedies, and human oversight.

Specifically, two public service systems are under review: the AI-assisted e-challan system operated by the Punjab Safe Cities Authority, and the Punjab Socio-Economic Registry (PSER) welfare assistance program.

The committee aims to scrutinize the accuracy of violation detection in the e-challan system, the process of human validation of violations, citizens' access to pertinent evidence, mechanisms for challenging decisions, and the handling of associated data. For the PSER program, the committee will look into registration and accessibility, data accuracy, eligibility assessments, explanations given for unfavorable decisions, corrections of erroneous information, and opportunities for reconsideration.

Crucially, the committee intends to establish the methodology used in determining eligibility, whether it's manual assessment, fixed rules, statistical scoring, machine learning, or a combination thereof. Furthermore, the committee will address the constitutional and legal ramifications of digital technologies, encompassing digital harms, deepfakes, threats to democratic trust, risks to women and vulnerable groups, data governance, and institutional capacity.