Global artificial intelligence (AI) governance must pair strict safety guardrails on leading-edge models with equitable access for developing nations to prevent new historical inequalities, United Nations (UN) Development Programme Administrator Alexander De Croo said Thursday.
Speaking to China Media Group in Beijing, De Croo sought to allay fears of widespread job destruction, stressing that AI is designed to augment human capabilities rather than simply eliminate workers.
He noted that managing the technology requires tackling two distinct global challenges, namely preventing high-end models from running out of control, and bridging the tech divide.
"Technology has always replaced certain jobs and this is not a new thing. It has been going on for more than hundreds of years with industrial revolutions and so on. We really live in highly technological times. At this moment, there are more people working than ever. At the moment, technology already plays a very important role. Here, with the artificial intelligence, we of course have a very powerful augmentation of our own intelligence," De Croo said.
He emphasized that the global discussion surrounding AI centers on two key debates, the first being the potential safety risks of leading-edge frontier models, which require strict manufacturer guardrails and international cooperation to prevent systems from running out of control.
"From our perspective, there are two different debates. One debate, and this is the one that we have seen heavily debated over the past months, is the danger of what we would call the frontier models, I mean the most leading-edge models. And it is clear that indeed there is a potential risk of some of those models being, as being said by some of the industry leaders, out of control, having their own mind, having their own priorities. And so it is of course the responsibility of producers to make sure that the necessary guardrails are there to make sure that, you know, it is not something that would go out of control. But, to enable that, you need international cooperation and you need a system," he said.
Addressing the second debate, De Croo underscored the critical need for equitable technology access, noting that many developing nations still lack the AI tools required to boost their development, making capacity building and governance essential.
"There is a second debate, and that second debate is one that actually that Chinese President Xi Jinping put forward three months ago, which is access. I mean, today, too many countries do not have access to the necessary technologies in artificial intelligence to accelerate development. And so that debate on fair access, that debate on diversifying access to technology are based on making sure that countries have the necessary capabilities, the necessary governance capabilities," he said.
De Croo’s comments echo President Xi Jinping’s address at the World AI Conference in Shanghai, where the Chinese leader urged international action to build Global South capacity and prevent a new historical AI divide.
UNDP chief urges global AI safety rules, equitable access for developing nations
UNDP chief urges global AI safety rules, equitable access for developing nations
