The policy must prioritise initiatives that are carefully adapted to the country's specific context
On August 14, 2025, the government endorsed the new National AI Policy, which has invited a lot of enthusiasm and positivity. By implementing this policy, the government hopes to achieve the digital transformation envisioned by the Digital Nepal Framework (DNF), which has identified and outlined seven sectors: smart infrastructure; agriculture; health; tourism; education; finance; governance, social protection, and lifecycle protection. Given their proliferating innovation and adoption, the government's effort to strategise and regulate the use of AI and other digital technologies is not just timely but praiseworthy. However, given Nepal's pre-existing digital divide and the sluggish pace of mitigating efforts, the implementation of the new AI policy not only faces significant challenges but also risks further exacerbating the digital divide, pushing rural and marginalised groups to further exclusion.
Before assessing their implementational feasibility, it is essential to recognise how Nepal's AI policy addresses inclusion. Notably, the policy's vision explicitly emphasises "inclusive" use of AI systems to "build a prosperous Nepal", signaling that the government acknowledges the importance of equitable access. To promote inclusion, the policy outlines targeted measures such as awareness, orientation, literacy, skill development, and capacity-building programmes at the federal, provincial, and local levels. These initiatives aim to accelerate AI adoption across diverse populations. Special focus on women and marginalised communities through training and workshops could help prevent the benefits of AI from being confined to urban, English-speaking groups.
The policy also promotes inclusive sectoral adoption. In education, it plans to introduce AI-related subjects in school curricula, support personalised and adaptive learning, and build skilled human resources through varied training programmes. Importantly, it encourages the use of local languages in AI development, making tools more accessible to non-English speakers. In agriculture, the policy envisions AI-driven innovations such as weather-based farming, smart irrigation, and e-agriculture markets. In healthcare, it aims to improve access and quality of services in remote areas. Furthermore, the policy seeks to enhance public service delivery through AI-enabled digitisation, which could increase transparency, improve grievance redressal, and simplify access to government services.
In essence, these are commendable goals that lay a solid foundation for inclusive AI development. However, Nepal's significant digital divide – marked by low digital literacy, inadequate infrastructure, and a shortage of skilled personnel – presents a serious implementation challenge, particularly in ensuring inclusivity.
To its credit, the policy does acknowledge key implementation challenges, including low digital and AI literacy, limited data availability and accessibility, poor digital infrastructure, and a shortage of skilled human resources. However, it falls short in incorporating these ground realities into its strategic planning and fails to propose context-specific, tailored solutions.
For instance, according to the Ministry of Education, Science and Technology, only 58.1% of government schools have ICT infrastructure, 43.1% have broadband connectivity, and just 25.1% have IT labs –many of which are not meaningfully accessible to students. Likewise, rural and marginalised populations remain poorly connected and digitally illiterate, with limited language proficiency to navigate even basic digital services such as e-health, e-governance, e-commerce, or digital banking, let alone engage with advanced AI systems.
Although the policy envisions legal, structural, and institutional measures to address these issues, many of its provisions appear impractical. For instance, while it proposes training teachers on AI, the government is already struggling to meet basic demands like salary hikes, making nationwide training unlikely. Teacher resistance, heavy workloads, and limited digital infrastructure further complicate this. Similarly, including AI-related subjects in school curricula or promoting AI literacy across all government levels seems overly ambitious, given the limited progress on digital equity and capacity building over the past two decades. Most schools and educators have yet to adequately respond to the broader digital revolution, and with the rapid pace of AI development, it is doubtful the government can act quickly or effectively. As such, the policy risks deepening the digital divide, benefiting only the already connected, urban, and digitally literate populations.
Nepal is not the only country facing challenges in AI development, adoption, and regulation. But for countries like Nepal, these challenges are particularly significant due to limited resources, infrastructure, and digital literacy. Therefore, the government's policy must prioritise initiatives and programmes that are carefully adapted to the country's specific context, addressing its unique social, economic, and technological realities. At the same time, it can observe what has worked in other countries.
Additionally, as the AI policy also encourages, the government should actively work alongside the private sector to address these foundational challenges. It must create a supportive environment where companies can confidently invest in, develop, and implement inclusive and ethical AI solutions that benefit all. This requires not only reducing bureaucratic red tapes but also providing the necessary infrastructure, resources, and additional incentives to enable meaningful innovation. Addressing these practical implementation challenges, especially the digital divide, will be crucial to ensuring that AI can truly contribute to Nepal's sustainable development.
Dahal is a software quality assurance engineer at Cedar Gate Technologies, Nepal
