Why the Greater Bay Area Is Uniquely Positioned for the Next Wave of AI Development

Martin Matsui of the Hong Kong Monetary Authority (HKMA) noted that Hong Kong's resilient banks, linked exchange rate system, and financial infrastructure make it a trusted testing ground for initiatives such as Stock Connect, Bond Connect, and Swap Connect . Yan Xuan of the Pacific Forum added that Hong Kong's common law system and respect for contractual rights underpin its value as an anchor for global companies . Institutional investors value "stability, trust and assets" above all — and Hong Kong delivers all three, creating the long-horizon capital environment that AI scale-up requires .

2. Energy Cost and Data Centre Infrastructure

China's energy advantage is structural. Cindy Chow of the Alibaba Hong Kong Entrepreneurs Fund highlighted that China's energy capacity is roughly eight times that of the US, at only 40% of the cost . China operates the world's largest power grid; its electricity generation surged more than the rest of the world combined from 2010 to 2024, and it produces over twice the electricity of the US . Data centres in China can be built at around 60% of US prices .

China has announced a roughly $295 billion (2 trillion yuan) national plan over five years to build interconnected AI computing hubs, with AI power demand driving rack densities from 20 kW to 50 kW by 2027 . Western data centres face aging grids and long delays for new power connections; China's grid is newer, larger, and can bring data centres online faster — a critical factor for AI compute returns . As one analysis noted, a one-year delay in operation would cost a 100-megawatt US data centre more than $500 million over its life cycle .

3. Open-Source AI Models

While China still trails the US in advanced chip production, it is "driving advancements in performance and in producing some of the world's top open-source models," Chow explained . Chinese AI labs including DeepSeek, Alibaba's Qwen, and Baidu's Ernie have released competitive open-weight models that dramatically lower the cost of entry for AI development across the GBA. Greater Bay Area AI Computing Tech secured over RMB 15 billion in new indicative orders for AI cloud services in the first half of 2026 alone, reflecting surging regional demand for AI compute .

4. A Digitally Native Population and Consumer Behaviour

The GBA's population is an asset that cannot be replicated. Chow described a "wealth of data generated over 20 years by a population that is extremely comfortable using and adapting to online technology" as something "that cannot be replicated elsewhere" . This gives GBA companies a decisive edge in training and refining AI models.

The second panel, moderated by Prof. Mandy Hu, discussed how AI is already transforming consumption patterns and brand approaches. Chris Tung of Shangri-La Group stressed that brands need "even stronger differentiation in today's AI era" because AI agents will shape consumer discovery . Sean Tam of Preface noted that AI agents can boost productivity for small and medium businesses across the GBA if they deploy AI strategically, adopt generative engine optimisation, and build distinct brand voices .

The Synthesis

The forum's core conclusion was that the GBA offers a rare combination no single hub can match: Hong Kong's stable, trusted capital platform to fund AI at scale; China's cheap energy and low-cost data centre construction to power it; world-class open-source models to build on; and a hyper-digitally native population that generates the real-world data and demand to refine and deploy AI rapidly. As the SCMP summary put it, "The full value of AI will only be realised via its continued development and implementation in the stable and fully integrated economic ecosystems and digitally receptive markets offered in regions like the GBA" .