Pakistan is set to develop its first indigenous Large Language Model (LLM) tailored specifically for its regional languages. The model will primarily focus on Urdu, with additional datasets for Pashto and Punjabi, addressing the shortage of AI resources for underrepresented languages.
By creating an AI model that supports these local languages, the initiative aims to bridge the digital divide and empower millions of people across the country to access advanced AI tools and services in their native tongues.
To bring this vision to life, VEON’s Pakistani subsidiary, Jazz, has entered into a strategic partnership with the National University of Sciences and Technology (NUST) and the National Information Technology Board (NITB). This collaboration is set to develop Pakistan’s first indigenous LLM designed to support underserved regional languages.
Empowering Local Communities through AI
This initiative is designed to meet the unique linguistic and cultural needs of Pakistan’s diverse population. The development of a local-language LLM will empower businesses, educators, healthcare professionals, and farmers by providing access to advanced AI tools in their native languages. The AI model is expected to be a game-changer for key sectors such as healthcare, education, and agriculture, where language barriers often hinder the effectiveness of existing technologies.
For companies and individuals in Pakistan, this development represents a transformative shift, with AI-powered tools tailored to local languages. Currently, many digital services globally are limited to widely spoken languages such as English, Mandarin, and Spanish, creating a significant gap in access and usability for people who speak less widely represented languages. With this initiative, VEON aims to reduce the digital divide by ensuring that people in Pakistan—especially those in sectors like healthcare, education, and agriculture—can interact with AI tools in their native languages.
The partnership between Jazz, NUST, and NITB marks a significant milestone for digital inclusion in Pakistan. The project is expected to not only improve user experience but also drive innovation and enhance productivity in several fields critical to Pakistan’s economic growth.
Key Highlights
- VEON’s Jazz has partnered with NUST and NITB to develop Pakistan’s first indigenous LLM.
- The LLM will focus on the national language Urdu, as well as underrepresented languages such as Pashto and Punjabi.
- The collaboration seeks to bridge the digital divide and promote digital inclusion in Pakistan.
- The partnership will empower individuals, businesses, and key sectors like healthcare, education, and agriculture through AI tools in native languages.
Closing the digital divide
Aamir Ibrahim, CEO of Jazz, stressed the importance of AI in transforming Pakistan’s future. “Innovation in AI and LLMs is critical for Pakistan’s future, and we are committed to developing a locally relevant environment that meets the unique needs of our communities,” Aamir Ibrahim stated.
“Through this partnership with NUST and NITB, we can help close the digital divide, empowering our teachers, doctors, and farmers by providing them with tools and information in their native languages. This collaboration is a national cause that will drive exponential progress in digital inclusion. In the coming months, we aim to have a functional model that will lay the foundation for a transformative AI framework in Pakistan,” he added.
“This partnership marks a pivotal moment for Pakistan’s digital future, enabling us to preserve and promote our linguistic and cultural heritage through advanced AI,” Pakistan’s Minister of IT & Telecommunication Shaza Fatima Khawaja said.
“By bringing together the strengths of academia, industry, and government, we’re not only building a platform that reflects Pakistan’s unique perspective but also positioning ourselves to lead in emerging technologies.” She said that the government will work to establish the necessary policies, resources, and frameworks to ensure this initiative’s success over the years.
