Since the release of OpenAI’s GPT-4 in March 2023, generative Artificial Intelligence (AI) has been perhaps the most talked about technological development globally.
As Laurel Miller, President of The Asia Foundation, observed last year, ‘AI technology is poised to have huge effects in developing countries.’
So this week we asked the experts: ‘old problems, new technologies: inequality, governance, fragility. What does AI mean for development?’
The international development sector should focus on identifying investments that can influence the future of AI technologies in away that prioritises the public interest, and especially the long-term interests of those in developing countries. To do this, program strategies –whether it be for programs mitigating harms, developing tools to solve specific problems, or creating policy frameworks for technology development and use – need to be based on careful thinking about long-term path dependencies.
In other words, we need to understand how the systems and solutions we support today will affect the ability of different communities and countries to maintain agency and benefit from opportunities in a world where AI becomes a foundational general-use technology.
This means using development programs to:
This approach tries to avoid the trap of rushing to engage an emerging technology, without paying sufficient attention to long-term sustainability, while also searching for alternatives to dependency on a technology that a handful of large corporations or states design, own, and control. Such dependency is not yet a forgone conclusion.
Toni is the Assistant Director – Technology, Policy and Innovation at The Asia Foundation. Toni is an expert on cyber policy and misinformation. She holds a Master’s degree in International Policy, specialising in Cyber Policy. At the Lab we love the Asia Foundation’s work considering the big disruptions coming down the line and how these will affect development.
As one of the founders of Dragonfly Thinking, a start-up dedicated to creating AI tools for unravelling complex problems and guiding better decision-making, I’m fascinated by what artificial intelligence can offer in reshaping policy-making. The age-old challenges we face in development—inequality, governance, fragility—are anything but new, but AI brings fresh eyes and tools to the task.
One of AI’s undeniable strengths is its capacity to weave vast amounts of data across disciplines into coherent insights. It holds the promise of helping governments and organisations grasp complex issues from multiple national and local perspectives, creating a rich tapestry of knowledge. It can also “bring up the rear” in capacity building, supporting governance efforts that might otherwise falter under the weight of limited resources.
But with this potential comes caution. Given the nature of the internet, many prominent AI models are disproportionately trained on English-language, US-centric data. This can mute the subtleties of local contexts, making the tools less responsive to the specific needs of regions like ours. There’s also the risk of policymakers leaning too heavily on AI, bypassing the necessary human oversight and the imperative to strengthen local expertise.
Despite these concerns, the chance to lighten the cognitive load on governments and enhance decision-making is too valuable to ignore. The key is ensuring we tread carefully, balancing the allure of AI with the need to amplify voices from the very regions Australian development policy is meant to assist.
Anthea is the CEO and founder of AI start-up Dragonfly Thinking. Anthea’s skills lie in teasing out the complexities and intersections in some of the biggest global issues. At the Lab, we can’t wait to see how Dragonfly Thinking evolves, and we’ll be keeping a keen eye on their work.
AI development and regulatory efforts are increasingly centralised in Western countries despite the impacts of – as well as inputs into – these technologies transcending borders. Western corporations at the forefront of AI innovation heavily depend on Global Majority countries for natural resources like tantalum, cobalt, and tungsten, which form the GPUs powering large-scale AI development. Additionally, human labour for data annotation and content moderation—critical for training AI systems—relies significantly on workers from East Africa, Latin America, and South and Southeast Asia.
The many basic development challenges still facing much of the Global Majority impede their ability to fully contribute to AI research and development. However, rising interest in adopting AI technologies and increasing participation in the global AI economy presents an opportunity for these countries to strengthen digital infrastructure and develop AI talent. This could improve economic outcomes and key development indicators, but will need strategic investments and synergistic cooperation from Western partners.
A number of risks from AI have already materialised. AI has amplified concerns around bias, disinformation, climate impacts, and economic inequality. Moreover, discourse on these topics often excludes non-Western perspectives. These concerns elevate a need for Global Majority countries to engage in comprehensive research and pursue robust governance efforts to mitigate the risks posed by AI technologies, while bolstering an equitable digital future.
Chinasa is a fellow in the Centre for Technology Innovation in the Governance Studies program at Brookings. She has deep expertise in computer science and AI governance. At the Lab we’re fascinated to see how her research explores the role of AI in the Global South and her articles are enjoyed by all our staff.
