I have long argued that governance is the key challenge associated with AI: we need rules, institutions, and accountability mechanisms that can mitigate the very real risks posed by this technology and guide its development so that it benefits the economy and society. So far, authorities have not been able to cope with this task.
Yes, attempts are being made to establish an effective AI governance system. In 2019, as a G20 Sherpa, I presented the OECD’s newly agreed-upon “AI Principles” for approval at the Osaka summitfor approval at the Osaka summit. A year later, the OECD launched the “Global Partnership on AI,” which now includes 46 countries. Although membership is ostensibly open, countries must first endorse the “OECD Principles” to join, even if they did not participate in their development.
In 2021, UNESCO adopted the “Recommendations on AI Ethics” (I oversaw their development). Although 194 countries, including China, have endorsed this document, they are implementing it unevenly, particularly with regard to key commitments on transparency, accountability, and the rule of law. Specialized agencies and UN organizations also have specific initiatives related to AI.
With the launch of ChatGPT in 2022, the general public was introduced to generative AI, which heightened the urgency of regulating AI. A year later, the G7 launched the “Hiroshima AI Process,” and the United Kingdom hosted the first AI Safety Summit at Bletchley Park.
Subsequent summits were hosted by South Korea in 2024, France in 2025, and India in 2026 (Switzerland will host the summit next year). One of the most promising outcomes of these summits was the creation of AI Safety Institutes (AISI)—initially in the United Kingdom and the United States. The international AISI network began operating in 2024.
These institutes focused on evaluating cutting-edge AI systems, improving AI risk assessments, and developing common safety standards.
The U.S. and China Are Taking Different Paths
However, under President Donald Trump, the focus of the U.S. AISI shifted to the narrower areas of national security and strategic competitiveness.
From the Trump administration’s perspective, serious AI regulation is impossible. According to this logic, regulation stifles innovation, and the U.S. cannot afford to fall behind China in the race for global technological leadership.
This same rhetoric of geopolitical competition prompted the U.S. to impose export controls on advanced chips and computing power, as well as to restrict foreign access to the most powerful AI models.
By some measures, the U.S. still clearly leads China. According to the “Stanford AI Index 2026,” U.S. private investment in AI totaled $286 billion in 2025, compared to just $12 billion in China. Fifty-nine notable AI models were developed in the U.S., compared to 35 in China.
However, the U.S. has good reason to be concerned. Chinese AI models—from DeepSeek to Kimi K3— are closing in on the capabilities of cutting-edge American systems, often at significantly lower costs. In addition, China is focusing on open-source models, which greatly facilitate access to AI for developing countries. And it is doing everything in its power to overtake the U.S. in AI development.
But the scope of competition is not limited to the AI market. Through WAICO, China aims to expand its geopolitical influence, including by becoming the power that sets the rules for AI for everyone else.
This aligns with the country’s overall strategy: since 2013, as part of the Belt and Road Initiative, it has allocated approximately $1.5 trillion to construction contracts and investments in partner countries. It is no coincidence that all of WAICO’s founding members (the organization will also focus on infrastructure investments and technology transfer) are participating in the Belt and Road Initiative (BRI).
China has already initiated the creation of the Asian Infrastructure Investment Bank, which, although it cannot yet compare in size to the World Bank or the Asian Development Bank, demonstrates China’s ability to create institutions—and its interest in doing so.
Meanwhile, the United States is doing the exact opposite: it is withdrawing from organizations that it itself helped create, including UNESCO, the World Health Organization, the United Nations Conference on Trade and Development (UNCTAD), and the United Nations Human Rights Council. The United States has also dismantled the U.S. Agency for International Development (USAID) and withdrawn from the Paris Climate Agreement.
The Trump administration’s decision to prioritize national competitiveness over the multilateral system is exacerbating the fragmentation of global AI governance.
The creation of WAICO takes this fragmentation to a new level by institutionalizing the Chinese initiative. The U.S. response was not long in coming. According to Reuters, the State Department is drafting a letter calling on all allies participating in U.S. initiatives (Pax Silica 2025 and the “Statement on AI Opportunities” 2026) to choose whose side they are on in the AI race.
Key issues remain unaddressed
None of these initiatives fully addresses the key issues that make people anxious about AI. These include models that create addiction and harm children and young people; discriminatory algorithms that perpetuate inequality; potentially existential risks of losing control; threats to national and cybersecurity; and upheavals in the labor market that we are barely able to assess, let alone address.
The only power attempting to address these issues is the European Union. Through the “AI Act,” the EU has introduced mandatory rules for the development and deployment of this technology, although it has postponed the entry into force of obligations for high-risk AI systems. Nevertheless, it supports developing countries in establishing principles that align AI with human rights.
However, Europe is not a leading player in the AI race, so it is difficult for it to give its regulatory initiatives global impact. Although Europe’s experience in other areas suggests that its regulations could still extend beyond European borders (especially if the number of AI-related incidents increases), this has not yet happened.
With the right regulation, AI will have the potential to help solve the world’s greatest problems. But for that to happen, the U.S. and China must adopt different priorities that are radically different from their current ones.
First and foremost, they will need to work together to ensure that AI serves the public good. For now, however, in a race that is supposedly being waged for the sake of humanity, humanity itself plays a secondary role.
Gabriela Ramos,
Co-Chair of the Working Group on Inequality and Social Financial Disclosure, former Deputy Director-General for Social and Human Sciences at UNESCO, where she oversaw the development of the “Recommendations on AI Ethics,” former Chief of Staff at the OECD, and a sherpa for the G20, G7, and APEC.
© Project Syndicate, 2026.
www.project-syndicate.org
Follow our updates
