Every day, new breakthroughs in artificial intelligence (AI) dominate headlines, urging policymakers to explore innovative ways to make the country’s governance system more effective, efficient and responsive.
In Germany, a self-financed company with just 10 employees has developed an AI-based animal care application that monitors foal births, tracks a mare’s health throughout pregnancy and provides real-time veterinary alerts. Another AI-enabled system monitors livestock health and ensures uninterrupted milk production. These self-learning AI systems are transforming traditional sectors into smart, data-driven industries, prompting policymakers to pursue AI integration even with limited financial resources.
Introducing AI into state institutions will, no doubt, enhance efficiency, provided pragmatic steps are taken at the right time and in the right direction. Traditional mechanisms may help sustain the system, but without adopting new technologies and methods, progress stagnates. Developed countries have already made significant progress in AI-driven transformation, while Pakistan has yet to harness AI to modernise public administration.
In Pakistan, internationally developed AI tools such as ChatGPT, DeepSeek and Gemini are already being used across government offices, universities, research institutions and media houses for drafting, translating, detecting errors, and interpreting policies and regulations. At the same time, the increasing reliance on foreign-developed AI applications raises an important concern: Is it prudent to submit official documents to AI platforms operated outside the country for analysis?
Advanced economies are pursuing AI sovereignty, while developing countries continue to rely on imported AI tools. The US and China have already established dominance in the global AI landscape. From the perspective of the realist school of international relations, AI technologies are not merely tools of innovation, but they can be strategically employed to advance political and economic interests and may emerge as geopolitical instruments in the future.
For Pakistan, the challenge is not only in adopting AI to improve governance but also in determining the extent to which the country should rely on foreign AI tools to perform critical government functions.
Unfortunately, the country has yet to develop the robust and reliable digital infrastructure required to adopt advanced AI technologies and reduce dependence on foreign AI applications. Building indigenous generative AI models, AI chatbots, and large language models (LLMs) requires substantial investment in GPUs, Tier IV data centres, reliable power infrastructure, advanced computing infrastructure, and high-speed networks.
Limited financial resources are not the only obstacle to AI adoption; several other challenges within the country have also impeded AI integration into Pakistan’s governance sector.
A lack of initiative remains a major impediment to Pakistan’s adoption of emerging technologies. Although the National AI Policy (2025) sets out an ambitious vision, its implementation has remained limited.
Institutional inertia impedes the implementation of modern administrative practices in public offices. Administrative officials are often reluctant to adopt new mechanisms and modern technologies.
Despite these challenges, Pakistan has begun introducing some indigenous AI systems in parliament and the Prime Minister’s Office. Similar initiatives should be expanded across government institutions to improve governance performance and reduce reliance on foreign-developed AI systems. AI systems for public services, complaint resolution, and tax administration can be developed domestically with limited financial resources at the departmental level, while international assistance may be sought where necessary for their design and development. Developing cost-effective AI systems within the country can significantly improve efficiency, transparency, and public trust in government.
The process of digitising records must be accelerated to create a conducive environment for AI applications in public offices. Establishing small data centres across government offices is practical and feasible.
Developing AI infrastructure in the country is imperative. Excessive reliance on other countries for AI-driven transformation is neither prudent nor sustainable. Protecting sensitive national data through domestic computing infrastructure is indispensable.
Tier III data centres are already operational in Pakistan. In the future, establishing Tier IV data centres for advanced AI applications is essential. Although this requires substantial financial investment, public-private partnerships should be encouraged to accelerate the country’s AI infrastructure development, safeguard national data, and reduce long-term dependence on foreign AI models.
The dream of an independent and responsible national AI ecosystem is achievable only if robust and prudent policies are designed and implemented in a timely manner. AI-driven public administration has the potential to be a real game-changer for Pakistan. If a small enterprise can develop an independent AI system to transform its business, there is no reason Pakistan cannot harness AI to transform the public sector and build an efficient, secure, and self-reliant state.
The decisions taken today will determine whether Pakistan becomes a creator of intelligence or merely a consumer of it.
The writer is an economics scholar at the Pakistan Institute of DevelopmentEconomics (PIDE).