BOAO, Hainan, March 27 (Xinhua) -- If the long queues snaking outside every AI-themed subforum and the buzz stirred up by robotic moderators and robot dancers at the venue are anything to go by, AI is absolutely the buzzword of this year's Boao Forum for Asia (BFA) annual conference.

Since its founding 25 years ago, the BFA -- held in Boao, in southern China's Hainan Province -- has championed international cooperation. This year, discussions have centered on how to foster global inclusion in artificial intelligence to prevent new divides, with a particular focus on China's role.

China's open-source AI ecosystem has materially lowered entry barriers for developers across the Global South, Sam Daws, senior advisor to the Oxford Martin AI Governance Initiative, Oxford University, noted in an interview with Xinhua on the sidelines of the BFA Annual Conference 2026, which ran from Tuesday to Friday.

"It provides low-cost and integrated full-stack solutions that enable sovereign AI development," Daws added.

"Without deliberate intervention, the concentration of AI infrastructure and talent in advanced economies will widen existing inequalities. The Global South risks becoming a consumer of AI technologies rather than a co-creator. This dynamic threatens to entrench technological dependency for generations," Daws cautioned.

His view was echoed by other speakers at the conference. Xue Lan, dean of the Institute for AI International Governance at Tsinghua University, believed that China's open-source AI models have made a significant contribution to universal access to artificial intelligence technologies worldwide.

As only a handful of countries have developed their own large language models, many nations are eager to adopt open-source models as a foundational layer -- a move that is particularly valuable for developing countries, where critical gaps in AI development persist, Xue added.

Some of these nations still lack basic infrastructure, such as reliable electricity and internet access. Through initiatives like the Belt and Road, China is well positioned to provide the necessary support and services to bridge these gaps, according to the professor.

Yu Xiaohui, dean of China Academy of Information and Communications Technology, supported the view, citing China's cooperation initiatives with ASEAN countries as examples, while joining an AI-themed panel discussion.

"I have been involved in several AI cooperation initiatives between China and ASEAN countries, and I have seen firsthand how eager regional leaders are to seize the opportunities of this technological wave. Their shared goal is clear: to keep pace and not be left behind," said Yu.

However, challenges lie ahead for countries in the region, particularly in developing models tailored to the diverse, smaller languages spoken across the region, Yu noted, adding that, on this front, China's AI companies have made significant contributions by offering fully open-source models such as DeepSeek, Kimi, and Qwen. These provide a strong foundation for countries to train models based on their own languages and build sovereign AI capabilities, he explained.

To address the challenge of computing power in Asia, Yu believed that strengthening regional connectivity and cooperation is essential to meeting growing demand. "By working together, we can better harness the innovation dividends of global AI development while safeguarding ourselves against risks and governance challenges," said Yu.

In light of risks and challenges posed by the burgeoning -- yet still relatively nascent -- digital technologies, experts at the forum called for a coordinated response.

Multilateral cooperation must harmonize safety and ethical standards across jurisdictions, according to Daws, who called for efforts to prevent regulatory fragmentation that could stifle innovation, and to establish equitable frameworks for governing computing resources and managing systemic risks.

Data from OpenRouter, a top global AI platform, showed that API calls, or user visits, to Chinese large models began surpassing U.S. levels in February, highlighting China's rising influence in the global AI ecosystem.

Chinese models occupy three of the top five spots globally by usage -- MiniMax's M2.5, Moonshot's Kimi K2.5, and DeepSeek's V3.2 -- mainly driven by their high cost-performance ratio, according to the platform.

China is actively promoting global AI governance. In 2023, China launched the Global AI Governance Initiative at the 78th UN General Assembly to enhance international cooperation on AI capacity building.