Pakistan is developing Buraq, a government-backed large language model designed to understand and generate Urdu and other regional languages, as part of a broader effort to build domestic artificial intelligence capabilities and keep Pakistani data within the country. Federal Minister for IT and Telecommunication Shaza Fatima Khawaja outlined the initiative at the Global AI and Digital Summit 2026 in Seoul, South Korea, during a World Bank panel on enabling digital government service delivery through digital public infrastructure and AI. Alongside the development of Buraq, the government is procuring and provisioning sovereign computing capacity to process Pakistani data within national borders and retain more of the economic value generated by digital services domestically. The initiative is intended to support the use of AI in public services while addressing concerns around data security, governance and local-language accessibility. By developing a model focused on Urdu and regional languages, Pakistan also aims to make AI tools more relevant to citizens who may not primarily use English-language digital services. The announcement places locally developed AI within a wider government programme to establish the digital systems needed for secure, accessible and scalable public services.
According to the minister, the government’s AI strategy rests on four foundational elements: digital identity, digital payments, trusted data exchange and data-governance safeguards. The first involves the National Database and Registration Authority (NADRA), whose identity system provides verified identification to more than 100 million Pakistanis. The government is working to extend this capability into digital identity services that can authenticate citizens online while recording and enforcing consent over the use of their information. The second element is Raast, the State Bank of Pakistan’s instant payment system, which can support the delivery of government payments, subsidies and social protection transfers. For example, an AI-enabled service could help identify households eligible for assistance, while the payment infrastructure provides a mechanism to deliver the funds. The third element is WASL, a national data exchange layer being developed under the World Bank-supported Digital Economy Enhancement Project. WASL is intended to enable government institutions to exchange information through authenticated, encrypted and logged transactions, with citizen consent built into the process. These systems are expected to provide the underlying infrastructure for digital government services rather than leaving individual departments to develop disconnected platforms.
The fourth foundation is the National Data Governance Policy, issued by the Pakistan Digital Authority, which establishes a national framework for classifying, sharing, protecting and publishing public-sector data. The policy is intended to provide consistent rules for government institutions while supporting accountability through audit trails and mechanisms for citizens to manage consent. The wider digital public infrastructure programme also includes plans for a citizen Super App and the National Open Data Ecosystem, known as NODE. These initiatives are intended to make digital services more accessible and provide lawful access to public data for researchers, startups and AI developers. The government’s approach reflects the view that an AI model alone cannot deliver reliable public services without accurate information, secure infrastructure and clear rules governing how data is used. Connecting identity verification, payment systems and data exchange could allow future AI applications to support administrative processes and service delivery across multiple government departments. However, the effectiveness of these systems will depend on their implementation, interoperability, security and the ability of institutions to follow common governance standards.
Khawaja emphasised at the Seoul summit that AI should be introduced after the foundational systems required to support it are established, warning that poorly prepared digital projects can remain isolated pilots that fail to scale or earn public trust. Buraq is intended to provide the AI layer of Pakistan’s broader digital infrastructure, while sovereign computing capacity is expected to help ensure that data processing takes place domestically. Keeping data within Pakistan could give public institutions greater control over sensitive information, although the practical benefits will depend on the infrastructure’s technical capabilities, security arrangements and governance. The government has not yet announced a public release date or detailed technical specifications for Buraq, and its performance will need to be assessed once further information and access become available. Its development nevertheless signals an effort to combine local-language AI with national data infrastructure and public-sector digital services. If implemented effectively, the programme could provide a foundation for more accessible Urdu and regional-language applications while supporting the secure exchange of information across government institutions. The broader challenge will be translating the planned infrastructure into dependable services that meet citizens’ needs and maintain transparency, privacy and human oversight.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
