{
  "id": 4706525,
  "title": "Beijing and Washington Can Build AI Safety Despite Mutual Distrust",
  "url": "https://urgent.news/2026/08/31/beijing-and-washington-can-build-ai-safety-despite-mutual-distrust",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-31T20:06:17.000Z",
  "source": {
    "name": "Foreign Policy",
    "slug": "foreign-policy",
    "url": "https://foreignpolicy.com/2026/08/31/china-us-safety-ai-artificial-intelligence-summit-regulation/"
  },
  "original_language": "en",
  "account": "Despite mutual distrust, Beijing and Washington have the potential to collaborate on AI safety, according to recent developments. The two global AI leaders have not held an official intergovernmental dialogue on artificial intelligence in over two years, despite the rapid pace of AI advancements and the emergence of frontier AI systems posing significant risks. The decision to initiate such a dialogue was made during U.S. President Donald Trump's visit to Beijing in May, a significant move since the last AI dialogue between the nations in May 2024. While the pace of AI development has accelerated, it is noteworthy that the two leading AI powers have gone more than two years without a formal conversation on this critical technology.\n\nSigns of convergence have emerged, with leaders of frontier AI labs in both countries openly discussing the importance of AI safety and security. Demis Hassabis, co-founder and CEO of Google DeepMind, has called for urgent action to address potential risks as AI approaches artificial general intelligence (AGI). Similarly, Chinese AI executives have publicly emphasized long-term development, safety evaluation, and responsible deployment of AI. Chinese AI firms are increasingly engaging with governance questions, reflecting a growing willingness to participate in discussions about AI safety and security.\n\nGovernments on both sides have also started to prioritize AI safety and security in their policy agendas. The Trump administration issued an executive order and presidential memorandum on AI security, integrating these issues within a national security framework. At the World Artificial Intelligence Conference in Shanghai, Chinese President Xi Jinping highlighted AI safety as a key priority, urging countries to strengthen risk awareness, ensure security and controllability, and prevent malicious use of AI. The WAIC chair's statement further emphasized maintaining AI safety as a fundamental principle and strengthening cooperation against AI abuse by terrorist organizations, extremist groups, and transnational criminal networks.\n\nHowever, despite these positive signs, growing strategic mistrust between the two nations threatens the success of the renewed dialogue. The launch of the World Artificial Intelligence Cooperation Organization (WAICO) in Shanghai has been perceived by some U.S. policy circles as China's response to Pax Silicon, reinforcing perceptions of escalating institutional competition over global AI governance. Simultaneously, technological competition has intensified, with concerns arising over potential knowledge transfer between Chinese and U.S. AI models. Sanctions and export controls have also been proposed by the U.S., while China has criticized such measures as unfair and hegemonic.\n\nMutual distrust has extended to AI security as well. The Trump administration has scrutinized Chinese AI products and infrastructure on cybersecurity grounds, restricting imports of certain Chinese AI hardware. China's National Vulnerability Database identified a potential backdoor vulnerability in Anthropic's Claude, leading Alibaba to prohibit its internal use. This reciprocal behavior could escalate tensions and jeopardize the dialogue that both governments have agreed to restart.\n\nThe renewed AI dialogue is occurring amidst growing policy contradictions. While Washington expresses interest in discussing shared frontier AI risks, it also views China's frontier AI development as a strategic threat, expanding sanctions, export controls, and other restrictive measures. Beijing, on the other hand, has been cautious in its approach, emphasizing the importance of cooperation, risk awareness, and ensuring AI remains under human control.",
  "summary": "Practical cooperation should exist alongside competition.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}