Sino-American AI cooperation is vital and feasible across most (albeit not all) areas. Hong Kong remains a distinctive part of China that possesses a greater range of freedoms and maneuvering room. Whilst geopolitical and domestic political forces have resulted in transformations to the city’s institutions and state-of-politics, the underlying virtues of the city remain clear: it can and should serve as the site where Sino-American engagement on critical issues central to humanity’s interests – such as climate change, biomedical hazards, and artificial intelligence – takes place.
The Importance of Sino-American AI Engagement – Revisited
Half a year ago, I wrote that it is vital for the U.S. and China to work together on AI regulation and safety. At San Francisco, Presidents Xi Jinping and Joe Biden affirmed the need for both powers to work together on artificial intelligence – settling upon an agreement for formal government-to-government discussions on the “risks of advanced AI systems and […] AI safety through U.S.-China government talks.” Whilst inter-governmental conversations are of course utile, what is equally pivotal is buy-in and involvement of non-governmental actors, with the richness of insights, connections, and expertise of academics advising regulators and policymakers (especially crucial when it comes to those working with the Cyberspace Administration of China, which helmed the country’s Global AI Governance Initiative), including those in the Chinese delegation who attended the Global AI Safety Summit last year. This explains the significance of broadening inter-governmental talks into Track-Two conversations, drawing upon entrepreneurs, academics, lawyers, and other stakeholders.
What would the scope of such engagement look like? For the sake of pragmatic feasibility and utile de-politicisation, there are bound to be areas that are off-limits: semiconductors and advanced manufacturing technologies are domains where Washington has clearly advocated a “small yard, high fence” strategy towards China. The White House’s primary emphasis remains containing, isolating, and limiting China’s ability to grow and develop in these sectors, and as such alignment remains elusive. Similarly, in particularly sensitive domains such as algorithms reflecting overarching military strategy and tactics, or the incorporation of AI into intelligence and counterintelligence – again, areas where the technology is employed to enhance cyber-operations as a means of ‘shaping,’ rather than ‘signalling’ in international conflicts – talk of Sino-American ‘alignment’ on AI would be pollyannish, and broadly inappropriate.
With that said, I submit that beyond these particular domains, there remain zones of interest convergence that should be taken advantage of – both sides of the Pacific do not want to be embroiled in a conflict where AI, and not human commanders or leaders, is steering decisions made on the battlefield; nor do they want to train AI models with adversarial prejudices towards the ‘other’ that result in over-amplification of risks, or over-simplification of the upshots and downsides of conflict. Finally, state-civilian interactions over AI regulations are a crucial domain in which both Beijing and Washington stand to learn from one another.
Why Hong Kong is Unique
Hong Kong is uniquely capable of serving as the forum and location for fruitful, in-depth engagement between China and the U.S. In December, t University of Hong Kong hosted a large conference aimed at “regulating Generative Artificial Intelligence,” bringing together legal experts, philosophers, and technologists from all corners of the world to tackle the incredibly salient question. The AI New Horizons 2023 symposium, hosted by the Hong Kong University of Science and Technology and the Greater Bay Area Association of Academicians, featured luminaries such as Yann LeCun, Harry Shum, and Pascale Fung, who convened at Asia Society Hong Kong to tackle some of the thorniest questions in AI regulation and development.
As a Special Administrative Region of China, Hong Kong is conveniently situated within the Greater Bay Area, a vivacious and propitiously world-leading cluster of innovation and technological commercialisation. It is a part of China, yet possesses significantly different – and more internationally oriented – laws concerning information and data flows, media platforms, and business and finance. Even amidst the geopolitical upheaval and fracas surrounding the city over the past couple of years, it can and should step up to the mantle of repairing and bolstering Sino-American relations, through providing the platform for both inter-governmental and government-to-citizen dialogues.
It is fair to say, this vision does not come without its challenges. OpenAI, for one, does not allow for Hong Kong IP addresses to use ChatGPT – placing the city in the same category as Mainland China and a handful of other territories and countries where the platform is unavailable. Some technology companies are concerned about their ability to operate in an unfettered manner within the city. There are ongoing concerns raised over Hong Kong’s ability to remain distinctive, vibrant, and dynamic. Yet I do not believe these questions, or the worries undergirding them, should be construed as dominant and obstinate impediments that render the city unfit for Sino-American AI conversations.
What Hong Kong Must Do
What can Hong Kong do? Here, Microsoft must be praised for its foresight and willingness to partner with Hong Kong universities in rolling out the ChatGPT tool to thousands of university faculty and students in the city. Yet more universal access would be needed in order to bring into the fold the rest of Hong Kong. As I have said, the most imminent and pressing priority is for relevant policymakers and its private sector to reach out to OpenAI and other leading AI companies in the U.S., and to position the city as a crucial ‘buffer zone’ between mainland China and the rest of the world, where in exchange for market access to the Greater Bay Area, they would be granted the unique opportunities to set up a research presence in the city. With one foot through the door, these firms can then pursue joint ventures and compare best practices with their mainland counterparts, thereby benefiting all parties involved.
The city must also strive to host a larger number of thematically specialised and outcome-oriented working group discussions between experts, academics, and even officials and bureaucrats in charge of AI regulation and policymaking, from China (both mainland and Hong Kong) and the U.S. Drawing upon the regionally unrivalled talents in the many world-class universities in the city, Hong Kong should press for its having a say and sway at the table when it comes to global AI governance and regulation. For one, given its long-standing advantages in the law, locally based think-tanks should perhaps focus on the intersection between AI, law, and the public policy consequences and implications of such interactions.
Hong Kong must remember that its most persistent and definitive strength lies with its openness, embracing of the international community, and willingness to seize upon the unknown and turn the unknown into real opportunities.
