A “widening” AI gap between advanced economies and developing countries has sparked calls to use the Asia-Pacific Economic Cooperation framework to develop an open, fairer and culturally sensitive governance system.
Although the technology is often hailed as a great leveller in an era of sluggish economic growth, a conference held in southern China last weekend heard that many Global South countries lacked the critical infrastructure and talent required to harness artificial intelligence effectively.
Speakers at the event warned there was a risk innovation could largely benefit a few countries and businesses, and proposed a range of solutions, including joint programmes and taking full advantage of Apec’s flexible structures.
Chung Chul, president of the Korea Economic Research Institute, said the economic gap was “widening, not shrinking” and an “inclusive” framework designed to tackle this problem “cannot be treated as a residual outcome. Instead, it should be something considered beforehand, not just after innovation.”
He said this principle should be fundamental to any form of technological cooperation and innovation, adding: “So the central policy question is not whether Apec should use AI but whether the region can scale innovation without widening the gap.”
Figures from the United Nations Conference on Trade and Development indicate the global AI landscape and its rule-making processes are heavily concentrated in the hands of a small group of nations and roughly 100 corporations.
Fewer than one third of developing countries have a national AI strategy, a figure that drops to just 12 per cent among the least developed nations.
The UN body also said 118 countries, mostly from the Global South, did not play a role in major international AI governance forums.
The event, hosted by the Institute of Public Policy at South China University of Technology, heard there was also a risk these countries would be marginalised culturally.
Hans d’Orville, a former assistant director general for strategic planning at Unesco, said AI systems were often trained on data sets that carried inherent linguistic and cultural biases.
“AI is not culturally neutral; it is reshaping culture at all levels and in all areas,” d’Orville said.
He warned that without responsible governance and cultural consideration, AI could reinforce existing biases and inequality, hinder cultural diversity and undermine local identities.
“The world urgently needs to ... introduce development-focused AI regulations to prevent the widening of the digital divide,” he said.
The conference also heard that Apec’s flexible nature could be used to establish an inclusive, open form of AI governance.
Chung proposed setting up joint training programmes, public-private partnerships and knowledge-sharing networks to help ensure all members of the bloc were able to integrate AI into their economies.
“The objective is not to make every economy develop identically, but to ensure that every economy has a credible pathway to participate in AI-empowered growth,” he said.
Chung also urged policymakers to make use of Apec’s pathfinder mechanism, which allows members to set preliminary rules to move forward on a specific policy issue without full participation from the rest of the bloc.
He also told the event that Apec’s regulatory sandboxes – which allow restrictions on members testing new technologies, products and business models to be relaxed – should be used to test cross-border data flows and AI governance.
He said these real-world cases could then serve as a guide when setting up regulatory frameworks.
Chung also called for measures to get a clearer picture of how well-prepared regulators, businesses – especially small and medium-sized enterprises – and workers were for the advent of AI.
“We shouldn’t just look at cutting-edge models or how many data centres we have, but evaluate whether people can use them widely,” he said. -- SOUTH CHINA MORNING POST
