Event: Discussion of AI Ethics with Policy Makers in Nepal
NAAMII organized a two-hour-long Focus Group Discussion (FGD) for Policymakers on 24 January 22 as part of our ongoing project to survey the state of “AI Ethics in Nepal”. The main goal of the event was to facilitate a discussion regarding the state of AI, AI ethics, and AI related policies in Nepal. The event was initially planned to be in-person but was adapted to virtual due to rising Covid-19 cases. There were a total of 8 policymakers and 2 team members from NAAMII. The participants were associated with Tribhuvan University (TU), Kathmandu University (KU), Pokhara University (PoU), Ministry of Education, Science and Technology (MoEST), Ministry of Communication and Information Technology (MoCIT), Nepal Academy of Science and Technology (NAST), University Grant Commission (UGC), Nepal Applied Mathematics and Informatics Institute for Research (NAAMII) and Policy Research Institute (a Government Think Tank).
The focus group was structured into four main topics of discussion – the opportunities and applications of AI in Nepal, ethical challenges with AI in the context of Nepal, AI/IT related policies in Nepal, and policies that are needed in Nepal.
In the first section, the participants discussed the classification of “Nepal” as a LMIC (Low- and Middle-Income Country) as well as an underdeveloped country. The participants suggested that AI could be useful to develop economically important sectors such as agriculture, tourism, healthcare, finance, and education and expressed dissatisfaction that very few of AI/IT companies are working in those areas. The AI experts among the participants mentioned that most AI/IT companies in Nepal are working as outsourcing companies for foreign multinationals, while a few startups are working on local AI products such as Nepali chatbots, recommendation engine, etc.
When asked about the ethical challenges for AI in Nepal, majority of participants suggested that we have to find gaps and loops in the current system, which we are currently unaware of. They talked about international guidelines such as the OECD guideline, EU guideline, UNESCO guideline and guidelines published by tech-companies such as Google and Facebook. They contrasted these guidelines with a lack of the AI specific policies in Nepal. The participants listed Nepal Health Research Council’s National Ethical Guidelines for Health Research 2019 and University Grand Commision’s Guideline for Research Misconduct as two main policy documents which guide research ethics in Nepal. They mentioned that these guidelines would be relevant for academic and independent AI research but would not cover corporate research or AI product development. Despite research ethics guidelines, the participants also pointed out a lack of AI ethics or Ethics literacy in formal curriculum. Some of the participants suggested including ethics as a subject in K-12 itself to create ethically minded human resources in the long run. However, other participants were not keen on this idea and reiterated a need for a national AI guideline.
In terms of AI specific policy, a participant mentioned the Science, Technology and Innovation Policy, 2019. This policy kept Artificial Intelligence as area of priorty and has a work plan for infrastructure, development and security. The participants in the FGD unanimously agreed that this policy was not sufficient and there is a need for separate policy and an AI review board. The AI review board (for research and technology audit), that may be an organization or a governmental body or infrastructure, which helps in making AI-related policy as well as auditing research and technology products. Some of the participants compared AI review board to the consumer rights board or human rights watch organizations. The participants also stressed the need to think about AI deployed systems globally from tech giants what are their ethical aspects and what their impact is on Nepal. The participants also discussed the inherent tradeoffs between building a national framework and regulatory mechanisms for accountable AI and reducing unnecessary bureaucracy and administrative costs for an active startup ecosystem. They concluded that a National Framework is needed but should be built with the above tradeoff in consideration and with involvement of stakeholders from different groups. Lastly, the participants also talked about lack of awareness and … in AI technology and the need for AI literacy and awareness programs targeted to different demographics.
With this discussion, the participants recommended that we should develop our AI norms and values and AI developing procedures contextualized to Nepal. They suggested that we should make a policy regarding Surveillance Ethics, Algorithmic Transparency, and the Intellectual Property of AI. We should consider making guidelines for AI companies to follow throughout the different stages AI development but these guidelines should be adaptive based on the use case. They suggested that we can work on ethical guidelines of responsible AI, by inter-relating health research ethical guidelines (NHRC, 2019), but they propose a National Ethical Guideline and add ethical guidelines to it.
Overall, this event was beneficial for us to understand the survey result and gaps between professionals and policymakers. Policymakers were little aware and concerned of the globally deployed AI system and its ethical aspects and impacts, but due to the exploitation of AI Talents by developed countries by outsourced business model operated in Nepal, we are lacking in developing our own AI products. They pointed out developing a governmental framework or institution, to build its policy and audit AI technology.
This project is generously funded by a grant from UNESCO Asia and Pacific Bureau.