Towards equalized AI: Insights from Oxford Professor Fu Xiaolan on democratized intelligence

Editor's Note: Alice Ho is a PhD candidate in International Development at the University of Oxford. The article reflects the author's views, and not necessarily those of CGTN.

In 2019, China issued the "Outline Development Plan for the Guangdong-Hong Kong-Macao Greater Bay Area (GBA)." This landmark document laid out comprehensive plans for the region's strategic positioning, development goals, and spatial layout, explicitly emphasizing the need to "promote deep integration of artificial intelligence (AI) with the real economy and vigorously advance manufacturing transformation and optimization." Six years on, the GBA has emerged as a global hub for AI innovation. With a population of 87 million and a GDP of $1.4 trillion, this region provides a unique ecosystem and testing ground for AI development.

What distinguishes the GBA is not just its focus on traditional AI metrics like algorithms, computing power, and data, but its emphasis on deep integration of AI across industries. From smart manufacturing in Shenzhen to fintech in Hong Kong and smart city initiatives in Guangzhou, the region demonstrates how AI can be democratized across various sectors and societies. The GBA's distinctive advantage lies in its combination of Guangdong's robust industrial foundation, Hong Kong and Macao's role as bridges between the Chinese mainland and the global community, and the region's rich pool of international talent and data resources.

As AI continues to reshape industries and societies worldwide, questions of access, equity, and inclusion have become increasingly central to global discussions about AI development. At the Oxford China Forum convened at Oxford University's Said Business School, a panel on "Emerging Technologies and Global Order: Imagining the Future of AI," brought together leading voices to discuss the pivotal transformations shaping our future. Professor Fu Xiaolan – founding director of Oxford's Technology and Management Centre for Development – shared a compelling vision of how nations, especially those in the Global South, might harness AI to drive more equitable development.

Fu opened with reflections on the rapid decline in AI costs, citing examples such as Chinese start-up DeepSeek. What was once an expensive, elite resource is becoming increasingly affordable and widespread – a pattern likened to the early diffusion of electricity. The falling costs of training and operating large language models mean that AI is no longer the exclusive domain of tech giants. Today, small enterprises, students, farmers, and governments in developing countries are beginning to tap into AI's transformative potential.

The shift signals the emergence of true AI democratization. Open-source models, affordable cloud platforms, and compact AI hardware are creating new opportunities for innovation far beyond the traditional power centers of Silicon Valley or Shenzhen. Instead of being a threat to humanity, AI – when guided wisely – has the potential to become a powerful tool for reducing inequality and enhancing participation in global innovation.

As with past technological revolutions, however, certain jobs will disappear, especially those centered on repetitive tasks. But consequently, new types of employment will emerge, such as AI trainers, digital agriculture consultants, and platform content curators. For these roles to take root, educational systems must evolve to prioritize adaptability, digital fluency, and the ability to collaborate with AI systems. Reskilling mid-career professionals – particularly those aged 40 to 60 – should be a national priority, as the speed of organizational change often outpaces policy responses.

On the other hand, "AI-reshaped comparative advantage" comes into effect as automation becomes more cost-effective and production begins to return to high-income economies, altering traditional assumptions about labor-based globalization. This could pose a challenge for developing nations that rely on low-cost labor as a competitive edge. Yet, it also opens new pathways: many AI-related tasks – from data labeling to system testing – require digital literacy more than capital-intensive infrastructure.

Drawing from her research in Bangladesh, Fu illustrated how digital technologies can empower marginalized populations. During the pandemic, mobile-based platforms powered by AI enabled women and youth in rural areas to access income and public visibility. A randomized experiment she conducted found that communities with access to digital tools experienced 30 percent less economic loss during COVID-19 compared to those without – highlighting the tangible value of digital inclusion.

This point echoes findings she shared at the 10th Multi-stakeholder Forum on Science, Technology and Innovation for the Sustainable Development Goals (STI Forum), hosted by the United Nations. Reflecting on the past decade of innovation, mobile phones, AI, and platforms have not only driven efficiency, but also provided new forms of employment and access to healthcare, education, and financial services. While digital technologies have enabled major progress in sectors like renewable energy and public health, challenges remain – especially in agriculture, where AI applications are still uneven, and in gender equity within science and technology fields. To address these disparities, she advocated for stronger youth participation, urging the UN and global institutions to create more space for university students and young innovators in shaping technology for sustainable development.

Returning to the question of infrastructure, AI accessibility must go hand in hand with investment in digital foundations, including fiber-optic networks, stable electricity, data centers, and 5G. Equally important is developing local talent: people who can not only use AI but also build and maintain the systems behind it. Initiatives like Digital Silk Road exemplify how infrastructure and capacity-building can be deployed together to spread the benefits of AI more evenly across regions.

China's potential in shaping this inclusive AI future featured prominently. As a leader in both AI development and digital ecosystem export through initiatives like the Belt and Road Initiative, China is well-positioned to help bridge the technological divide – particularly by sharing platforms, training, and research collaborations with Global South partners. However, such engagement must prioritize mutual benefit.

To make AI a tool for sustainable and inclusive growth, countries must invest in foundational infrastructure, expand access to education, and develop flexible, forward-thinking policies for workforce transition. Above all, global cooperation – especially between developed and developing nations – will be crucial to ensuring AI's promise is shared broadly.

The AI era is no longer a distant possibility; it is an unfolding reality. As Fu noted, the question is not whether AI will reshape our world, but whether we will build the systems, skills, and solidarity needed to shape that transformation for the better.

Successful AI development requires more than technological prowess – it demands an integrated approach that combines policy support, diverse application scenarios, and cross-border collaboration. As the GBA continues to bridge technological capabilities with practical applications across different sectors and communities, it provides an instructive model for how regions worldwide can work towards more inclusive and sustainable AI development. The GBA's evolution demonstrates that the future of AI lies not in isolated innovation hubs, but in connected ecosystems that can translate technological advances into broader societal benefits.