(Yicai) Aug. 4 -- Bridging the global divide in artificial intelligence and enabling more countries to share in its benefits is a pressing challenge for the international community, according to Song Haitao, dean and president of the Shanghai Artificial Intelligence Research Institute.

Fresh from visiting Brazil, a BRICS member country where he visited two major national science labs and top universities, Song threw himself into organizing the Emerging Markets and Developing Countries Forum on Bridging the AI Divide and the opening ceremony of AIM Global CoE, both held yesterday at the 2025 World Artificial Intelligence Conference in Shanghai.

“BRICS countries now attach great importance to emerging technologies, from government leaders to various institutions at all levels,” Song said in an interview with Yicai. “As we enter the fourth and fifth industrial revolutions, there’s a growing consensus that AI is a pivotal tool and general-purpose technology driving this wave of transformation.”

In Brazil, he visited the National Laboratory for Scientific Computing and the Brazilian Center for Physics Research. “They’re investing in areas like quantum communication and optical computing,” Song said. Rio de Janeiro’s city government is also leveraging AI and digital technologies for public services such as urban safety monitoring and healthcare management, he added.

“Brazil's substantial investment and development in AI and high-performance computing was beyond my expectations,” Song said.

But “objectively speaking, the US and China lead globally in AI technology,” he noted, outlining three key reasons why they have pulled ahead. First is talent accumulation. China, with its vast population, has a strong foundation in mathematics and sciences, while the United States, as a hub for global talent, attracts top minds.

Second is scale of investment. China and the US together invest far more in AI research and infrastructure than all other countries combined, providing a solid base for breakthroughs. And the third reason is a complete industrial chain. AI requires integrated hardware and software ecosystems, and China’s manufacturing prowess offers a significant advantage.

“In the past 30 to 40 years, we followed global tech trends,” he pointed out. “Now in AI, we’re racing alongside the US, driven by technology.”

Despite gaps in core technologies, emerging markets are carving out their own paths. Brazil, for instance, has capabilities in electronic design automation and chip design, albeit on a limited scale, Song said.

“Over 90 percent of countries lack a complete chip supply chain. Brazil’s retention of this R&D capacity is no small feat," he said, noting that this foundation paves the way for future leaps. In applications, Brazil’s traditional strengths in agriculture and livestock provide unique opportunities for AI adoption.

China’s Approach 

To address the global AI divide,China has proposed the “AI From China, Benefits the World” initiative, offering tailored solutions to foster inclusive development.  Song revealed that in collaboration with partner countries such as Brazil, they wish that China can help build "sovereign AI" and "sovereign models." 

"First, we respect each country’s cultural and historical context,” he said. “Second, for sectors involving public safety and strategic resources, we prioritize adapting technologies, respecting local strategic and security concerns.”

This cooperative approach rests on three pillars: openness, inclusivity, and diversity. "First, we use open-source models, sharing technological benefits transparently while ensuring security. Second, inclusivity means respecting the independence of different countries, cultures, and languages in technology applications. Third, diversity entails honoring each nation’s unique needs," Song explained.

The development of AI governance frameworks demands a departure from traditional technology regulation, embracing a new paradigm that integrates technological ethics, civilizational diversity, and the provision of global public goods. A robust AI governance system must be multidimensional, emphasizing technology for good, ethical oversight, and global collaboration.

Challenges remain, however. Invisible barriers often impede cross-regional collaboration. As Song, an expert in the field, notes, "Current international scientific research cooperation faces new dynamics and challenges. The narrowing of certain technical exchange channels has created significant hurdles for joint research." He highlights that the efficiency of global innovation networks, rather than predictable technological advancements, will likely determine future AI breakthroughs. "Real progress often stems from open knowledge flows, requiring all stakeholders to apply professional expertise to build resilient collaborative frameworks," he adds.

China-U.S. cooperation is pivotal to global scientific advancement, particularly in AI. Both nations, as AI powerhouses, bring unique strengths to the table. By fostering exchanges while adhering to the principle of technology for good, they can achieve a synergistic effect, yielding solutions to global challenges that surpass the sum of their parts. Despite short-term obstacles, such as technology blockades, Song remains optimistic. "Complete isolation is unsustainable due to global talent flows, knowledge exchange, and free trade systems," he asserts.

Yet he remains optimistic for the long term. “Short-term tech blockades create temporary obstacles, but complete isolation is unsustainable due to global talent flows, knowledge exchange, and free trade systems,” Song said.

Accordingly, China pursues multiple cooperative avenues through the United Nations and other international bodies to build AI inclusive capacity to forge technical‑standard ecosystems, as well as through international tech sector exchanges, he said. By promoting extensive technical exchanges, China aims to drive the development of standards, applications, and ecosystems for emerging technologies, fostering a collaborative global environment for AI innovation.

Shanghai Model

As a scholar turned industry promoter, Song has clear views on the commercialisation of AI. He said the technology is moving from consumer entertainment into the “deep water” of industrial manufacturing, a sector that is set to be the center of commercial growth and investment value over the next five to 15 years.

Song is particularly bullish on intelligent robotics. “Smart robots will flood industrial, consumer, household, and community settings, just as personal computers, smart phones and cars did -- billions globally,” he said. “The integration of robots into society and homes in the future represents a massive market.”

The Shanghai Artificial Intelligence Research Institute has incubated in excess of 30 companies, raising more than CNY6 billion (USD835 million), primarily through market-driven funds from industrial companies such as battery giant Contemporary Amperex Technology and venture capital firms like Sequoia Capital.

He believes that AI entrepreneurs are moving from "technological innovation" to "scenario application" and are gradually moving toward the "paradigm definition era." For AI entrepreneurs, Song offers three key tips: focus on real‑world needs rather than chasing fads; build long‑term capacity, including stable teams and sustained cashflows; and focus early on global expansion.

As a key hub for China’s AI development, Song’ institute is pioneering a full-chain innovation model from research to industrialization. Through deep cooperation with the United Nations, Shanghai Cooperation Organization, BRICS countries like Brazil and India, this model is being pushed globally to form a network for inclusive AI development. Upcoming collaborations include visits from the Rio de Janeiro municipal government and Brazilian university leaders to Shanghai to sign agreements advancing high-performance computing and AI applications, he said.

On the first day of the WAIC, The Global Alliance on AI for Industry and Manufacturing’s Center of Excellence Innovation Network Initiative was launched.This pioneering international institution, jointly established by the United Nations Industrial Development Organization (UNIDO), China’s Ministry of Industry and Information Technology, and the Shanghai Municipal People’s Government, marks the first UN framework dedicated to fostering AI cooperation.

That followed the 2024 joint inauguration by UNIDO, China’s Ministry of Industry and Information Technology, and the Shanghai government, of the world’s first Global Industrial AI Alliance Center of Excellence. Located in Shanghai, the initiative aims to further bridge the digital divide between nations and deepen the exchange of ideas and practices.

“My personal vision is to become a key driving force and participant in making AI benefits universal worldwide,” Song said. From promoting open‑source models to harmonizing technical standards, from fostering international talent to diversifying application scenarios, China is taking concrete steps to advance inclusive global AI development.

Editor: Tom Litting