Unlocking AI potential in Asia’s developing nations

The advent of AI is rapidly changing the global social-economic landscape. As we eagerly await the next breakthroughs to be presented by AI agents, physical AI and more, one-third of the global population are still waiting to be connected online. According to United Nations, out of the top 100 high-performance computing (HPC) clusters in the world capable of training large AI models, none is hosted in a developing country. In a world polarised by economic and technological divides, will the deployment of AI exacerbate inequality and poverty?

While acknowledging the reality of uneven distribution of resources worldwide, Stephanie Hung King-chung (BBA 1991, DBA student), Chief Information Officer and Director General of Information Technology Department at the Asian Development Bank (ADB), shares her optimism for what AI can bring to the developing world.

Revolutionising developmental work from within

Stephanie has been spearheading the initiative to integrate Generative AI models into its operations and analytical work at ADB, harnessing the vast knowledge and experience accumulated in the past 50-plus years.

“Whether we are building a road, a hospital, water facility, or a solar plant, we’ve produced a lot of reports and accumulated a wealth of knowledge over the years across the region,” says Stephanie. Now, using big data, the new AI models can efficiently extract useful insights from the massive information bank. Such data has enabled us to conduct more accurate and detailed assessments, providing valuable insights into policy and project recommendations to their member countries.

With the support of AI, data across different departments and functions can be synchronised effortlessly. This capability has significantly improved cross-domain analyses and risk assessments related to climate change, natural disasters and financial market volatility, enhancing the effectiveness of solution deployment.

From roads to solar plants, Stephanie highlights how ADB leverages AI to extract valuable insights and knowledge from vast data (Photo courtesy: Shutterstock)

Tiering transformation in the field

With enhanced capacity to produce powerful insights, how are they translated into actionable changes on the ground?

Stephanie sees AI as a catalyst for unprecedented growth in developing nations, though pace and scope vary from country to country. A tailored, multi-tiered approach is necessary to address each country’s specific needs based on digital maturity.

For countries like Uzbekistan and some Central Asian nations, digital platforms and solutions can be used to enhance decision making on road constructions, ensuring their international safety standards and limiting carbon emissions. Others are devising digital and AI strategies to boost specific sector growth, such as tourism, healthcare and agriculture, while tackling basic infrastructural challenges like clean water, sanitation and energy.

“We will provide financial support and technical assistance to help them build those capabilities,” explains Stephanie.

Stephanie believes that a multi-tiered approach should be adopted to help Asian developing countries to address their specific developmental needs using AI based on their level of digital maturity

Building blocks for progress

On the other hand, however, establishing a robust foundation of nationwide infrastructure remains a priority in some countries. For instance, many remote Pacific Island nations lack undersea fiber optic cables, leaving approximately 80 per cent of their populations without internet connectivity. This divide highlights the stark disparity in access to information and communication technologies, which directly impacts socioeconomic opportunities.

“How do we enable inclusive and affordable internet access to these populations, and then also provide digital literacy training and cyber security protection? There’s quite a lot of work ahead for us to do as a development bank.” Stephanie notes.

Once the hardware is in place, it is also critical that the local workforce is adequately trained to put the wheels in motion. In this light, AI-enabled education could offer a beacon of hope.

Over the past 40 year, digital technology that has demonstrated immense potential to transform education. More recently, breakthroughs in AI have made personalised learning accessible even to remote communities with limited teaching resources, paving the way for a revolutionary shift in education systems and literacy enhancement.

The digital divide highlights the stark disparity in access to information and communication technologies, which directly impacts socioeconomic opportunities (Image generated by DALL-E)

Steering the AI bias

While the foundation of digital infrastructure is essential, the deployment of AI must go beyond technical capabilities. As AI continues to evolve and intertwin with humanity – from daily life to national decisions – critical thinking and ethical judgement are of paramount importance to ensure AI operates fairly and inclusively.

Stephanie illustrates, “For example in rural development, when a lending officer relied solely on a bank’s data to make lending decisions, only 20 per cent of loans were extended to women, as men typically had longer credit histories and held more jobs.” She adds, “However, if we increase the lending to women, particularly small household farmers, by 40 per cent, they can buy more seeds for the next crops. With the return on these investments, their credit history will improve. In this scenario, human intervention plays a key role in enhancing fairness in future data,” Stephanie says.

In response to these challenges, Stephanie is developing a “responsible use AI framework” for ADB. This framework addresses issues such as privacy, security, safety, reliability, accountability, transparency and human agency. As AI continues to advance, it holds the potential to transform societies, but this transformation must be guided by human values and principles to achieve betterment for all.