The world's largest gathering dedicated to AI opened in Shanghai on July 17, at a moment when both the promise and the dangers of the transformative technology have never been more apparent.

The four-day World AI Conference (WAIC) and the High-Level Meeting on Global AI Governance came just 10 days after the United Nations' first Global Dialogue on AI Governance in Geneva, Switzerland, where the International Scientific Panel on AI, a 40-member body created by the UN General Assembly last August, presented its inaugural global assessment.

Yet one question dominated both events, a question that has increasingly preoccupied economists since data emerged alongside oil and gold as one of the world's most valuable strategic resources: Who will write the rules of the AI age?

The World Cup: AI's living laboratory

This question was particularly prominent during the FIFA World Cup Final between Spain and defending champion Argentina at MetLife Stadium on July 19, which concluded the largest tournament in football history, featuring 48 national teams and 104 matches hosted across Canada, Mexico and the United States. An economic impact assessment commissioned by the organization estimates that the tournament will generate $80.1 billion in global economic output.

Behind the spectacle, however, operates an enormous digital ecosystem. A growing number of specialists argue that the tournament's broader "digital halo," including digital advertising, betting markets, algorithm-driven audience engagement and AI-powered influence over public opinion, could ultimately represent an economic ecosystem worth more than half a trillion dollars. While this estimate exceeds conventional industry calculations, it reflects an emerging reality: Big data—not the trophy—has become the event's most valuable prize.

Experts increasingly warn that algorithms are no longer merely analyzing football matches. They are determining what billions of people watch, how they think, what they feel and even what they bet on, yet this unprecedented influence operates with remarkably little international oversight.

An industry divided against itself

The debate over AI governance is no longer confined to governments; it has reached the heart of Silicon Valley itself.

In 2024, American entrepreneur Elon Musk, an early investor in DeepMind before its acquisition by Google and a co-founder of the U.S. startup OpenAI, filed a lawsuit against OpenAI and its partner Microsoft, alleging that the companies had abandoned OpenAI's original mission of developing safe, open-source AI in favor of commercial profit.

Although a California court dismissed the case on procedural grounds in May, the dispute exposed a profound divide within the AI industry between those who prioritize ethical development and those who place commercial incentives first.

Meanwhile, the risks are no longer theoretical.

Researchers found in 2019 that 96 percent of deepfake videos circulating online consisted of non-consensual pornographic content.

By 2023, that number had climbed to approximately 96,000 videos, representing an increase of 550 percent, while other estimates suggested the total exceeded half a million videos by mid-2024.

The Internet Watch Foundation, a UK-based watchdog, subsequently described 2025 as the worst year in Internet history for child sexual abuse material after AI-generated videos depicting such crimes increased by more than 260-fold within a single year.

Another study found that so-called "nudify apps" had been downloaded more than 705 million times, while researchers at the Oxford Internet Institute identified approximately 35,000 digital undressing tools, collectively downloaded over 25 million times since late 2022.

Within an 11-day period alone, the Center for Countering Digital Hate, a nonprofit with offices in London and Washington, D.C. that fights online misinformation, reported that Grok, the AI assistant developed for Musk's X platform, generated roughly 23,000 pieces of content associated with child sexual exploitation.

These technologies extend far beyond synthetic pornography.

They are increasingly deployed to produce hate speech, orchestrate industrial-scale disinformation campaigns and manipulate political narratives.

Observers argue that these operations disproportionately target conflict zones and countries across the Global South through misleading datasets amplified by major search engines and social media platforms.

The open-source counteroffensive

At the same time, the global balance of power in the AI landscape is shifting rapidly and Shanghai has become one of its principal battlegrounds.

China's open-source AI movement accelerated in January 2025 with the launch of DeepSeek R1, which demonstrated performance comparable to leading American frontier models at a fraction of the development cost. The announcement contributed to an estimated $600-billion decline in AI chip maker Nvidia's market capitalization during a single trading session, the largest one-day market value loss in U.S. stock market history.

Eighteen months later, the transformation has become even more striking.

Chinese tech giant Alibaba's Qwen family of models has surpassed one billion downloads and developers have created more than 200,000 derivative models, making it the largest AI ecosystem on AI collaboration platform Hugging Face.

Since August 2025, Chinese open-source models have overtaken their American counterparts in cumulative downloads on the platform, occupying seven of the positions in the top 10 by the end of the year and accounting for nearly 30 percent of global usage, compared with barely 1 percent in late 2024.

A report published in March by the U.S.-China Economic and Security Review Commission found that 80 percent of American AI startups now rely on Chinese open-source models in developing their applications.

Shanghai showcased this industrial momentum with more than 300 global product launches during the WAIC, including Huawei's next-generation high-performance computing clusters and the first smartphones designed to function as autonomous AI agents.

China's AI adoption rate among major industrial enterprises has already exceeded 30 percent, while domestic production of humanoid robots is expected to reach 100,000 units this year.

The Global South's AI agenda

Beijing is seeking to complement technological leadership with new international institutions.

On July 16, 29 countries signed an agreement in Shanghai on establishing the World AI Cooperation Organization (WAICO), an independent intergovernmental international organization headquartered in Shanghai.

At last year's WAIC, Chinese Premier Li Qiang proposed establishing the organization, which is open to all countries and has reducing the digital and AI divide between developed and developing nations among its central objectives.

The Asia Center for Studies and Translation, of which I am founder and director, has fought tirelessly since 2017 to champion sovereign search engines, indigenous social media platforms and self-reliant AI ecosystems designed exclusively for the Global South, and every recent breakthrough in the AI space irrefutably proves how forward-thinking and vital this long-held vision truly is.

What we are witnessing today is definitive proof: China has demonstrated that meaningful competition is possible. Qwen, DeepSeek, GLM and Kimi are open-source, freely accessible models competing directly with America's leading AI systems. Yet the Global South still needs its own digital platforms, including free AI-powered video-generation tools, to protect its research, cultural heritage and intellectual property. Most importantly, technologies possessing such extraordinary power and equally profound risks must ultimately be governed by ethical and legal standards that advance the development of our societies rather than simply maximizing corporate profits.

The author is an Egyptian economist and director of the Asia Center for Studies and Translation. He is also a non-resident senior research fellow at Vienna International Institute for Middle East Studies

Copyedited by G.P. Wilson

Comments to dingying@cicgamericas.com