Pakistan has introduced a governance framework under the National Data Governance Policy that will require government departments using Artificial Intelligence or automated decision-making systems in high-impact or rights-related matters to follow stricter standards of transparency, accountability, and human oversight, as the country prepares for broader adoption of Artificial Intelligence across the public sector.
Under the policy, government agencies deploying Artificial Intelligence applications must ensure their systems are explainable and auditable, carry out risk assessments before deployment, and maintain records explaining how the systems function and reach their outputs. Citizens affected by significant automated decisions will be entitled to meaningful human review where appropriate, establishing a right of recourse that did not previously exist in Pakistan’s public sector digital services framework. Public bodies must also implement safeguards against algorithmic bias and discrimination and embed privacy protections throughout the development and deployment lifecycle of any Artificial Intelligence-powered service.
To formalise oversight structures, the framework introduces a dedicated governance profile covering Artificial Intelligence, automated decision-making, emerging technologies, and spatial data. Supporting standards accompanying the policy will introduce requirements for Artificial Intelligence model accountability, explainability, performance monitoring, model drift management, and the mandatory registration of high-risk Artificial Intelligence systems before they can be deployed within government services. The model drift management requirement is particularly significant, as it acknowledges that Artificial Intelligence systems can degrade in accuracy or fairness over time as the data environment they operate in changes, creating ongoing obligations for departments rather than a one-time deployment approval.
The Pakistan Digital Authority will oversee implementation through compliance assessments, technical standards, and periodic audits, giving the authority a formal enforcement role in the Artificial Intelligence governance space that complements its broader digital transformation mandate. The policy frames its objective as encouraging the responsible use of Artificial Intelligence in public administration while ensuring that government systems remain transparent, secure, and accountable to the citizens they serve, a balance that will be tested in practice as Artificial Intelligence adoption across federal and provincial institutions accelerates in the years ahead.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
