Seventy years after a group of young scholars first coined the term “artificial intelligence” at the 1956 Dartmouth Conference, the 2026 World Artificial Intelligence Conference (WAIC) concluded in Shanghai this month at what may prove an equally defining moment. In his keynote address, President Xi Jinping articulated a vision of AI developing in a direction that is “upward and good” and that benefits humanity, putting forward four proposals: upholding openness and win-win cooperation to drive innovation; strengthening risk awareness to ensure safety and controllability; encouraging inclusiveness to promote mutual learning among civilizations; and advocating solidarity to improve global governance.
The timing could not be more consequential. As AI evolves at breakneck speed, a dangerous bifurcation is emerging. Driven by geopolitical friction and mounting data walls, the global tech ecosystem is decoupling into distinct Eastern and Western spheres. Against this backdrop, China is working to build a fair and equitable global AI governance framework that benefits all of humanity, particularly the Global South. For the Hong Kong Special Administrative Region, this national blueprint presents an unprecedented opportunity — and a critical question: What is the HKSAR’s strategic role?
Hong Kong’s destiny in the AI era cannot be relegated to that of a passive technology adopter or a mere compliance enforcer. Under the “one country, two systems” framework and with a globally trusted common law system, the city possesses strategic value no other city can replicate. Each of the four proposals articulated at WAIC maps directly onto an advantage Hong Kong already holds — and onto its most potent strategic engine, the Northern Metropolis.
Northern Metropolis as a ‘data customs’ hub
The first proposal emphasizes open-source collaboration and cooperative sharing to let AI empower every industry. Historically, discussions surrounding Hong Kong’s AI development have fixated on computing power — such as the 5,000 petaflops milestone highlighted in the 2026-27 Budget. However, while hardware can be purchased, institutional advantages must be cultivated.
Hong Kong cannot afford to be a mere spectator. We already have the assets: the data facility clusters planned for Sandy Ridge, the AI Supercomputing Centre at Cyberport, the robust privacy ordinances, and the immense spatial potential of the Northern Metropolis. What is required now is the strategic boldness to connect them. By transforming the Northern Metropolis into the world’s premier sandbox for AI governance and cross-border data flow, Hong Kong will not merely play its part in the symphony. It will help write the score — establishing a “Hong Kong standard” for the safe, inclusive, and prosperous AI era that WAIC 2026 envisioned. The blueprint has been drawn; it is time for Hong Kong to act
The Northern Metropolis — particularly the Hetao Shenzhen-Hong Kong Science and Technology Innovation Cooperation Zone, and the San Tin Technopole — must not be merely a real estate project. The government should strategically position it as a cross-border data free-trade zone or a “data customs” hub: a secure, closed-loop environment where high-value, anonymized industry data from the Chinese mainland — spanning healthcare, smart mobility, and finance — can flow across the border in an orderly, regulated manner, while multinational tech giants bring their global datasets into the same zone. With the city’s world-class universities partnering with tech giants across the Guangdong-Hong Kong-Macao Greater Bay Area, international researchers and mainland enterprises could codevelop algorithms, share computing power, and commercialize AI applications for global markets — precisely the kind of AI moving from the digital into the physical world that the WAIC vision anticipates.
The common law ‘anchor of trust’
The second proposal stresses that AI must remain a trustworthy tool under human control, while opposing the overgeneralization of national security concepts in the AI field. This is arguably where Hong Kong’s greatest strategic value lies. The rapid proliferation of AI brings both inherent and derivative risks — from data privacy breaches and deepfakes to algorithmic bias — and when investors and tech giants choose a base for research and development, their primary considerations are legal certainty and robust intellectual property protection.
As the only common law jurisdiction within China, the HKSAR can act as the nation’s regulatory sandbox and legal firewall. While the Digital Policy Office’s Ethical Artificial Intelligence Framework and the Hong Kong Generative Artificial Intelligence Technical and Application Guideline mark a commendable first step, the city must evolve from regulator to proactive enabler. By building upon the generative AI sandboxes already pioneered by the Hong Kong Monetary Authority and Cyberport, we can expand this model into the Northern Metropolis to encompass high-stakes sectors such as medical tech and autonomous driving. Within these sandboxes, enterprises could test product vulnerabilities using innovative “AI vs AI” stress tests, receiving early regulatory feedback before products reach the public — an approach that builds exactly the risk-monitoring and early-warning capabilities the WAIC vision calls for, while reducing compliance costs for enterprises. International companies can develop and test their models here; mainland firms expanding overseas can use Hong Kong as a base to ensure compliance with international ethical and data security standards.
Training AI that understands the world
The third proposal cautioned that AI must not erode the diversity of world civilizations or the cultural uniqueness of nations, calling instead for a flourishing “garden of civilizations” where each blossoms in its own way and all thrive together. Yet the global AI landscape remains dominated by models trained on single-region data, prone to algorithmic bias and a homogenization of cultural values.
Hong Kong, as a cosmopolitan city where East meets West, is the ideal testing ground for culturally inclusive AI. Its bilingual environment and diverse demographic makeup — combined with the controlled data convergence a Northern Metropolis sandbox would enable — could allow developers to train large language models that genuinely understand Asian cultures, languages, and contexts. By developing such models here, we can help the nation export technology that respects the cultural uniqueness of Belt and Road Initiative partners.
From superconnector to super-enabler
Finally, Xi called for true multilateralism to bridge the digital and intelligence divide and to avoid new historical injustices in the AI field. He announced that over the next five years, China will provide 5,000 AI training placements for developing countries, build international AI application cooperation centers for the Association of Southeast Asian Nations (ASEAN), the Arab League, the African Union, the Community of Latin American and Caribbean States, the Shanghai Cooperation Organization, and BRICS countries, and roll out the “Mazu” intelligent meteorological early-warning system in 30 countries.
Hong Kong has always been a superconnector, but its role must now evolve into that of a super-enabler for the Global South. It should proactively bid to host or complement the ASEAN-facing international AI application cooperation center within the Northern Metropolis, welcoming scholars, policymakers, and entrepreneurs from across the developing world. As a premier international financial center, Hong Kong can also provide the green and tech financing that Global South partners need to build their own AI infrastructure, data centers, and smart city projects.
From solo to symphony
As President Xi observed, AI development should not be a solo performed by any single country, but a symphony of global cooperation. The establishment of the World AI Cooperation Organization in Shanghai — a vision conceived just one year ago, now realized in response to the call of the Global South — is a testament to that commitment. This year marks the opening year of the nation’s 15th Five-Year Plan (2026-30), a blueprint that balances development with security and offers the international community a list of opportunities. Hong Kong cannot afford to be a mere spectator. We already have the assets: the data facility clusters planned for Sandy Ridge, the AI Supercomputing Centre at Cyberport, the robust privacy ordinances, and the immense spatial potential of the Northern Metropolis. What is required now is the strategic boldness to connect them. By transforming the Northern Metropolis into the world’s premier sandbox for AI governance and cross-border data flow, Hong Kong will not merely play its part in the symphony. It will help write the score — establishing a “Hong Kong standard” for the safe, inclusive, and prosperous AI era that WAIC 2026 envisioned. The blueprint has been drawn; it is time for Hong Kong to act.
The author is a member of the Legislative Council of the Hong Kong Special Administrative Region and a Hong Kong deputy to the National People’s Congress.
The views do not necessarily reflect those of China Daily.
