How to maximize the benefits of AI while curtailing potential risks?

Editor's note: Fabrizio Hochschild is UN Under Secretary-General and Special Adviser for the UN Secretary-General on digital cooperation. The following is the transcript of his keynote speech at the Tsinghua University International AI Cooperation and Governance Forum 2020, edited for clarity and brevity. The article reflects the author's views and not necessarily those of CGTN.

AI technologies are being used in everything from commercial services to public services, in areas as diverse as education, health care, infrastructure, dating apps, and much more. We are moving towards inhabiting ever smarter cities where AI will utilize public data to identify the most efficient distribution of resources relating to transportation services, utilities and waste management.

Our private lives are also increasingly influenced by AI applications, the onset of the COVID-19 pandemic has only accelerated and accentuated the use of AI. AI has been instrumental in tracking the disease, predicting its evolution, advancing diagnostic treatments, and vaccine research development.

As 2020 comes to a close, the most pronounced proliferation of the use of AI is probably in the health care sector. Accenture has predicted by 2021, the applications of AI and health care are expected to grow by no less than 40 percent.

The pandemic has also heightened our dependence on the predictive capacities of AI systems. On the one hand, this allows us to better understand and cope with the evolution of COVID as well as with climate change and other threatening global phenomena. On the other hand, this predictive capacity has also proven unreliable in unexpected circumstances, making us vulnerable to systemic errors. This has been salient in e-commerce predictions, which were unable to anticipate human behavior during the pandemic, which led to significant disruption to global supply and demand. AI's predictive capacity can also be used to manipulate consumers and voters to fuel polarization. As we saw with the Cambridge Analytical case and has documented in the Netflix documentary, the social dilemma.

The limitations and dangers of AI are also becoming clearer. Malicious actors can use AI for ever more sophisticated cyber attacks, while lethal autonomous weapons can be used for crimes with no easily traceable perpetrator. AI can be used for mass surveillance, whether for reasons of political suppression or for commercial exploitation. Harm can also come from unintentional misuse, like discriminatory algorithms based on bias data that amplify unequal access to jobs, to justice, or to finance.

All this shows how we have yet to fully understand the societal implications of the ever-expanding use of artificial intelligence.

Our ability to maximize the benefits of AI while curtailing potential risks is harmed by the growing fragmentation in the digital space. We are seeing the geopolitical fault line between major powers with technology emerging as the new battleground. Superpower rivalry and frictions are made worse by the deepening digital divide between the North and the developing south.

A total of 3.6 billion people, mostly in developing countries, remain unconnected to the internet, and for them, the benefits of AI are a distant dream. The global south is lagging far behind in patents, intellectual property, and expertise relating to AI. The others are dependent on tools and expertise from more developed countries, as well as vulnerable to data exploitation practices that prevent their ownership over their own data or even visibility of how their data is used.

To address this complicated, challenging global landscape, we urgently need global leadership and global multi-stakeholder cooperation at the highest levels. No single country or company can design comprehensive and anticipatory guidelines to manage the rise of AI and its ripple effect around the globe. We must come together to create an international viable cooperation and governance framework for AI.

However, there are a number of challenges related to realizing such an objective, the first is the digital divide I spoke of before. Developed countries with extensive, high-speed broadband networks are rapidly adopting AI applications, far outpacing the rate at which it's happening in developing countries.

Secondly, many existing initiatives on artificial intelligence lack any representation and engagement from the global south.

The third challenge is one of the big data, the fuel for AI. Any standardization of data sharing must encounter challenges of inaccurate and incomplete data in developing countries without excluding them. Moreover, we need to ensure greater diversity and data sets to help prevent social and cultural biases from being perpetuated and amplified by AI systems. We also need international principles that underpin how citizens' data is utilized, stored, and shared so as to protect fundamental human rights, like the right to privacy.

All of this means we need flexible and innovative forms of global AI cooperation and governance that prioritize the responsible, transparent use of AI and the privacy and protection of AI uses.

A number of initiatives like the OECD (AI) Policy Observatory, the global Partnership on AI (PAI), and the international congress for the governance of AI are actively working to support such international AI governance efforts, but much work remains, particularly in ensuring greater inclusivity, an adequate global representation in AI decision making process. This is where the United Nations can play an important role in bringing all concerned governments, the private sector, civil society, academia, the technological community to the same table to work together.

The United Nations Secretary General António Guterres has made clear that how we address the challenges of the digital world is one of the key issues of our time. He thus launched a road map for digital cooperation, which laid out a vision on key digital issues, such as universal connectivity, digital human rights, and digital inclusion. In his road map, he specifically highlights AI as an area that needs greater global steerage.

This is part of important work being done by the broader UN family in this domain, with UNESCO working on global AI ethic standards, ITU on building capacity on AI for good, UNICEF on AI for children. The Secretary General has also called for a ban on the use of lethal, autonomous weapons.

Of all the emerging technologies, artificial intelligence stands alone as the one with the greatest potential to empower but also to disrupt. This is why the stakes for international cooperation in this area are the highest. The fact that AI technology applications advance faster than not only normative and regulatory frameworks but also faster than our ability to understand their impact on us underscores the urgency of this call.