Nepal needs a credible plan to regulate AI

What a gulf exists between the Federal Government’s aims to promote Nepal as an international hub for information technology and the existing state of play of regulating Artificial Intelligence in the country. Regardless of official declarations, policies and budget speeches focused on harnessing the IT sector, the country’s quest to, first, make sense, then regulate and finally leverage the unfolding AI revolution is still in its infancy.

Amid this scenario, the very first official policy document on AI, technically a concept paper coming from the Federal Government, might offer the much-needed blueprint upon which new policies and legislations can be drafted. If Nepal really wants to attract investors in the field of IT, then it needs to really get it right the way it is going to regulate artificial intelligence. The fact that the country is already a late comer in understanding how to regulate the former could be seen as an advantage if the government acts swiftly.

It needs to leap, jumping with decisiveness by quickly taking advantage of and internalizing the learning and experiences from those trailblazing nations that, in the last few years, have been breaking ground in terms of AI regulations. To do so, it is equally indispensable to work in partnership with experts from the civil society and the international community.

To review the latest developments and take stock of what is happening in this complex but fascinating area of policy making, I got in touch with two persons involved in the discussions around AI, Santosh Sigdel and Aakriti Kharel, executive director at Digital Rights Nepal and digital media specialist at UNESCO Nepal Office, respectively.

“The adoption of the Concept Paper on the Application and Practice of Artificial Intelligence by the government of Nepal is a welcome development. Key aspects of this paper include the prioritization of developing AI laws and policies, such as the AI Policy Framework, National AI Strategy, Data Protection Framework, AI Governance Structure, AI standardization, and the promotion of AI literacy” shares Sigdel to me via e-mail.

Indeed, a very comprehensive framework is what is required. Let’s not forget that AI could offer humanity some of the best ways to leverage progress for the common good but, we know very well, that the opposite is also very realistically possible. Unchecked and unregulated, AI can become a devastatingly effective tool against humanity.

Sigdel is crystal clear on the downsides of the AI revolution and we should not live under the illusion that a still developing nation like Nepal will be immune from them. “As AI technologies become more integrated into daily life, it is crucial for citizens to understand these technologies to benefit from their advantages and mitigate potential risks. AI intersects with human rights in significant ways, impacting not only digital rights but also other fundamental rights. AI systems can collect and analyze vast amounts of personal data, raising privacy concerns,” he shared.

Think about discrimination and inequalities, two elements that are still very much embedded in society. “AI has the potential to either mitigate or exacerbate existing inequalities and biases. If not carefully designed and implemented, AI systems can perpetuate and even amplify biases present in training data, leading to discriminatory outcomes in critical areas such as law enforcement and access to public services,” Sigdel adds.

The risks are so high that the United Nations has been at the forefront, pressing for a global discussion around the ethics of AI. The Secretary-General of the UN, Antonio Guterres, even established an AI Advisory Board in October last year and the upcoming Summit of the Future in September, probably Guterres’ most ambitious undertaking since taking the helm of the UN since 2017, will try to hammer out an agreement on a Global Digital Compact that also will include aspects related to AI governance.

Within the UN system, UNESCO has been at the forefront of the conversation. “Nepal recognizes the significant impact of AI and is actively working on its ethical development” Kharel shares with me. “The UNESCO Recommendation on AI Ethics, adopted globally in 2021, serves as a vital guide for Nepal. This framework emphasizes human rights, transparency, fairness and human oversight in AI systems—values that align with Nepal’s focus on data privacy and ethical practices” she adds.

Positively, the Concept Note that was recently launched in a major event in Kathmandu is detailed enough to potentially pave the way for the Federal Government and the Parliament to take real and tangible actions.

According to the note, Nepal might have its own specialized AI agency. “The consideration of establishing a dedicated and specialized nodal agency to promote the use, application, regulation and governance of AI, as well as the encouragement of self-regulation, is positive,” Sigdel explains in his response. Will such a recommendation be acted upon?

We also need to ensure that any future policy making process related to AI is open and inclusive approach. These are two essential features for AI’s development. Sigdel strongly supports this view. “The process of AI policy-making should be open, transparent, consultative and participatory. The government should ensure the participation of all major stakeholders, including civil society, in the AI policy-making process.” “Civil society organizations should also be vigilant of the policy-making processes to ensure that diverse perspectives, including those of marginalized and vulnerable groups, are considered. This helps in addressing potential biases and ensuring inclusive AI practices,” he explains.

Kharel also guided me through what UNESCO has been doing to help the complex policy making cycle as much open as possible. “In Nov 2023, UNESCO and the Ministry of Communications and Information Technology (MoCIT) together with the Kathmandu University and Digital Rights Nepal organized the Multistakeholder Dialogue on AI Governance.” “The event brought together government officials, civil society, academia, and tech leaders to discuss AI policy. While Nepal’s specific AI regulations are still in progress, discussions highlight an application-based approach that prioritizes data privacy, responsible development and protection for vulnerable populations,” she adds.

While there is an urgency for Nepal to follow through the policy-related recommendations of the Concept Note, we cannot ignore the basics, among other concerns: Digital and media literacy.

“One key recommendation on AI Ethics is for member-states to invest in and promote digital and media literacy to strengthen critical thinking and understanding of AI systems, thereby countering misinformation and hate speech. UNESCO recognizes the risk of AI spreading misinformation, especially in Nepal,” the UNESCO expert tells me.

So, at the end of the day, it should not only be about regulations of the AI sector in order to generate incomes for the country. No doubt that a strong policy framework based on the best policies available, starting from the EU AI Act and then adjusted to local context, is going to be instrumental to truly make Nepal an IT hub. But the implications of AI’s use and spread are much broader and certainly cross-cutting along the whole spectrum of policy-making.  

“UNESCO has been at the forefront conducting awareness-raising dialogues on information integrity, engaging with youth and civil society on media and information literacy, enhancing capacity of female journalists on digital safety, collaborated with local governments to integrate media literacy in school curricula, trained judges on international standards of human rights and freedom of expression in AI contexts”. The private sector has a self-interest in helping Nepal come up with a strong AI policy framework with clear guardrails. 

Kharel explained that UNESCO also encourages tech companies to adopt ethical AI guidelines to prevent hate speech and misinformation.

Can Nepal win the AI challenge? The stakes are very high. This is something that a resolute and determined Prime Minister like KP Oli can tackle head-on if he wants. Yet Oli also needs to thread through it carefully. The broader society must be engaged and involved. Oli would be dead wrong if he acted too fast and too furious.

Amid a myriad of problems, will AI become one of the PM’s priorities? Will he bother to listen to relevant experts? Will he embrace a fast-paced yet balanced approach to regulate and harness its positive potential for the development of Nepal, while minimizing the risks associated with it?