'Without Deliberate Correction, AI Will Give Old Discrimination New Speed and Authority' - UN Deputy Secretary-General at the AI and Human Development Regional Convening
Welcome to this convening on Artificial Intelligence and Human Development. Thank you for joining us in Xi’an.
I would like to begin with my sincere thanks to the Government of China, Zhao Gang, Governor of Shaanxi Province, and to Liu Bin, Assistant Minister of Foreign Affairs, for their warm welcome and partnership.
Allow me to also thank our Resident Coordinator, Stephen Jackson and his team, for their dedication and hard work to help bring this gathering to life.
There is good reason why the Asian leg of this conversation begins here. China is helping set the global pace on artificial intelligence.
Only last month, the Secretary-General of the United Nations, António Guterres, joined President Xi Jinping at the World Artificial Intelligence Conference in Shanghai. When we spoke afterwards, he told me how astonished he had been by what he saw: not only the sophistication of the technology, but the speed and scale at which it is already being deployed.
He was equally clear about China’s wider contribution: launching the Global AI Governance Initiative and helping advance the first United Nations General Assembly resolution on AI capacity-building.
He called China’s open-source strategy “an enormous global success”.
So we want to learn from China’s experience. We want to understand what this technology looks like once it leaves the laboratory and enters people’s lives. And we want China’s help in ensuring that the capability being built here advances human development far beyond its own borders.
Ladies and gentlemen,
When it comes to artificial intelligence, the Secretary-General has insisted on a simple principle: technology must serve people.
Under his leadership, the United Nations has worked to bring every country into that conversation. The Global Digital Compact has given countries a framework for cooperation. The Global Dialogue on AI Governance and the Independent International Scientific Panel will help the world understand this technology and govern it.
But this series of convenings addresses the question that neither innovation nor governance can answer alone: what is AI doing to human beings, to our relationships and to the societies we share in all of this?
We have heard voices on AI from across Africa and Latin America, and for this third leg, we bring Asia’s histories, experiences and values into a global conversation.
We chose Xi’an as the setting for this convening with care.
This city was once a starting point for the Silk Roads, the network historian Peter Frankopan has called the world’s “central nervous system”.
It was from here that, in the seventh century, Xuanzang set out in search of Buddhist texts. He returned with hundreds of manuscripts and spent years translating them.
Knowledge crossed borders. Human understanding grew.
Could artificial intelligence become a new central nervous system for our age?
If so, who will be connected to it?
In Cairo and Mexico City, our discussions kept returning to one small word.
“We.”
We say that “we” must govern AI and decide what it is for.
But who is “we”?
That question has particular force in Asia.
More than half of humanity lives in this region. Some Asian countries - like China - are building the world’s most advanced AI systems while millions of people are still waiting for reliable electricity or an internet connection. At the current pace, Asia and the Pacific will miss 88 per cent of its measurable Sustainable Development Goal targets by 2030.
Women are still vastly underrepresented in AI research. They are missing from the teams that build artificial intelligence, and from the data on which it learns, and the result is that algorithms are turning yesterday’s discrimination into tomorrow’s automated decision.
Feed an algorithm decades of unequal hiring practices, and it will learn that a man is the stronger candidate. Train it on a world in which women have had less access to wealth, and it will interpret that exclusion as financial risk. Build it with medical data centered around male bodies, and women’s symptoms will be missed.
Without deliberate correction, artificial intelligence will just give old discrimination new speed and authority.
Young people are facing a version of the same imbalance: we are preparing them for a labour market, and we have no idea what that market is going to look like. They need the tools and skills to work with these systems, with the AI beings and the AI jobs that will make up the future, but we are simply not giving them what they need.
Right now, the generation that will live with this technology longest has the least say in how it is built.
So all of this begs the questions: who will help write this next chapter of human development? Who will be written into somebody else’s version of it?
The discussions in Cairo and Mexico City also warned us against romanticising ‘humanity’. More human input to AI systems will not automatically produce more humane technology.
Because humanity is capable of both extraordinary courage and terrible cruelty.
We have used knowledge to cure disease, and to perfect weapons of mass destruction.
We have declared every person equal in dignity, and yet we violate that promise every day.
And AI is learning from all of it: we are teaching machines from the full archive of human behaviour before we have agreed what lessons they should draw from it.
We must decide what in our humanity deserves a place in the future we are building.
Asia’s traditions can help us do that.
Here in China, the Confucian principle of Ren teaches that our humanity is formed through our obligations to others.
In Indonesia, Gotong Royong gives that idea practical form. When work is too heavy for one pair of hands, people join their effort: my well-being cannot be separated from yours.
We owe it to one another to anchor that same sense of shared responsibility in artificial intelligence.
Across Asia, cooperation is beginning to translate those principles into practice.
China’s AI Capacity-Building Action Plan for Good and for All extends that ambition to the Global South, calling for equal opportunities under rules countries help shape together.
ASEAN’s Guide on AI Governance and Ethics and its Responsible AI Roadmap establish common ground between countries with different systems and levels of capacity.
The APEC Artificial Intelligence Initiative makes meaningful participation and capacity-building explicit goals.
These efforts point towards a future in which technological leadership builds capability beyond its own borders and smaller economies help set the terms.
Xi’an once connected civilizations through routes along which knowledge passed from one people to another.
Today, Asia can help build a new kind of connection, one in which technological power is shared and intelligence remains answerable to humanity.
That would be leadership worthy of this region.
That would be progress worthy of the name.
Ladies and gentlemen,
We began this journey in Cairo and continued it to Mexico City. Now we come to Xi’an.
From here, we travel to the Vatican, before bringing these perspectives together at the United Nations in New York this September.
But New York will not be the end of this journey.
In 2027, we will take the conversation closer to where people live, through country-led and community-based convenings around the world, in the lead-up to the SDG Summit.
Each gathering brings another region’s history, experience and values into the conversation. Each one challenges what we thought we knew and reveals questions the previous gathering did not.
Each of you has a unique perspective. You see something that others may miss.
Keep that in mind today.
We are looking for curiosity and truth, not consensus.
Give us your questions as much as your answers.
Challenge one another - play the devil’s advocate.
Ask what Ren demands of those who design, govern, or even use AI.
Come with an open mind, and come with humility, because no one in any country, company or laboratory has the full answer.
We do not have all the answers yet.
That is precisely why we are here.
The future will contain more intelligence than any civilization before us could have imagined.
Whether that intelligence brings greater wisdom and serves human dignity remains our decision.
Thank you.
These remarks were delivered at the AI and Human Development Regional Convening.
