Microsoft co-founder Bill Gates is seeking a meeting with Chinese President Xi Jinping later this year to discuss international measures for controlling some of the most dangerous applications of artificial intelligence. Gates told Reuters he believes China could support restrictions on the release of potentially dangerous AI models if the United States takes the lead, while also proposing international monitoring of AI systems capable of assisting biological attacks or creating harmful molecules.
Gates is also calling for a new international organisation dedicated to AI governance, drawing comparisons with institutions created to manage nuclear inspections, aviation safety and environmental risks. His concerns have intensified following recent demonstrations of AI systems carrying out sophisticated cyberattacks, alongside wider worries about job displacement and psychological dependence on AI. His foundation plans to spend $200 billion by 2045, with AI becoming an increasingly important part of its work.
🚨WTH. Bill Gates now wants governments to tax AI when it replaces human labor and says society could reserve up to 40% of today’s jobs for humans.
Gates is seeking a meeting with Xi Jinping, tentatively in November, to push for global AI rules. He believes China might restrict… pic.twitter.com/JwY1ynRoTH
— Choblin (@choblin29) August 26, 2026
The proposed conversation comes as governments are already taking very different approaches to AI regulation. The European Union’s AI Act entered its main application phase on August 2, 2026, giving the bloc a formal enforcement structure for regulating high-risk and general-purpose AI.
AI regulation is becoming a global security issue, not just a technology debate
The biggest change is the speed at which AI capabilities are moving beyond conventional software risks. Stanford’s 2026 AI Index recorded 362 documented AI-related incidents in 2025, up from 233 in 2024. That represents a rise of about 55% in one year. Policymakers have to consider what happens when increasingly capable systems are connected to real-world tools, sensitive data, biological research or cybersecurity infrastructure.
The economic stakes make AI regulation a security issue as much as an economic one. US private AI investment reached $285.9 billion in 2025, compared with $12.4 billion in China, according to Stanford. With so much money and strategic interest behind AI development, governments have little incentive to slow down while competitors move ahead. But the risks are no longer limited to technology or jobs. More capable AI systems could be used to develop harmful biological materials, support cyberattacks or strengthen other forms of national security threats.
Those risks also do not stop at national borders. An AI model developed in one country can be accessed, copied or adapted somewhere else, particularly when its technology is made openly available. That makes it difficult for any government to manage the most dangerous applications on its own. This is where Gates’ proposal takes on a broader security dimension: if AI can create risks that cross borders, governments may eventually need shared rules for identifying and limiting those risks before they become international crises.
You may also like: BitGo Expands MiCA-Compliant Crypto Services Across Europe
Can the US, China and Europe actually agree on AI rules?
The three major AI powers are not starting from the same regulatory position. The EU has built a binding, risk-based framework, while the US has relied more heavily on a combination of federal policy, executive action and state-level rules. China has taken a more state-directed approach, combining AI regulation with industrial policy and controls over how certain AI systems are developed and deployed. That makes a single global rulebook difficult because each side is trying to protect different economic and strategic interests.
Still, there are signs that governments are willing to cooperate where the risks become too serious to manage alone. In July, 29 countries agreed to establish the World AI Cooperation Organization, while more than 100 countries and international organisations participated in China’s World AI Conference. The UN has also begun a Global Dialogue on AI Governance. These efforts suggest that international coordination is developing, but mostly around principles and cooperation rather than enforceable limits on the most powerful AI systems.
The harder issue is whether Washington and Beijing can agree when AI has become a strategic competition. The US attracted more than 23 times China’s private AI investment in 2025, according to Stanford, while both countries are competing for computing power, advanced chips, talent and influence over global AI standards. Any agreement that restricts powerful models could therefore be viewed through the lens of technological advantage, not simply safety.
That is where Gates’ proposed discussion with Xi becomes relevant. A global agreement does not necessarily have to begin with rules covering every AI application. The more realistic starting point could be a narrow set of risks that neither country wants to see escalate, such as AI-assisted biological attacks, autonomous cyber operations and the uncontrolled release of highly capable models. If the US and China can establish common safeguards in those areas, Europe and other countries would have a stronger basis for joining them. If they cannot, AI regulation is likely to remain divided along geopolitical lines even as the technology becomes increasingly global.
Meanwhile, the United States and China are planning high-level talks on artificial intelligence in September as the two countries confront the risks created by increasingly powerful AI systems, five people familiar with the matter told Reuters.
Enjoyed this? Bookmark DeFi Planet, explore related topics, and follow us on Twitter, LinkedIn, Facebook, Instagram, Threads, and CoinMarketCap Community for seamless access to high-quality industry insights
Take control of your crypto portfolio with DEFI PLANET PRO, DeFi Planet’s suite of analytics tools.
