US and China agree AI shouldn’t control nuclear weapons

The first-ever agreement between the United States and China on AI principles marks a significant diplomatic milestone, particularly regarding the control of nuclear weapons and military applications of artificial intelligence.

Key meeting details: Presidents Joe Biden and Xi Jinping convened in Peru for comprehensive discussions that yielded important agreements on artificial intelligence governance and nuclear weapons control.

• Both leaders engaged in what their respective governments described as “constructive” dialogue about artificial intelligence development and implementation
• The meeting produced parallel statements from the White House and China’s Ministry of Foreign Affairs, highlighting areas of agreement
• Both nations endorsed the principle that humans must retain control over nuclear weapons decisions

AI military applications: The discussions revealed shared concerns about the intersection of artificial intelligence and military capabilities.

• The White House emphasized both nations’ commitment to cautious development of military AI applications
• China’s alleged use of Meta’s open-source Llama AI model for military intelligence has raised concerns about dual-use AI technology
• The case of DJI, currently suing the US for its placement on a “Chinese military companies” list, illustrates ongoing tensions over civilian versus military technology applications

Cybersecurity tensions: Despite the cooperative tone on AI governance, significant disagreements persist regarding cyber operations and security.

• Biden expressed “deep concerns” about Chinese cyberattacks targeting US critical infrastructure
• China categorically denied conducting cyberattacks, asserting itself as a victim of such activities
• Security researchers have linked Chinese hacking groups like Salt Typhoon to major breaches at US telecommunications companies, targeting high-profile individuals including former President-elect Donald Trump
• Reports indicate Chinese hackers have targeted OpenAI staff through spear-phishing campaigns

Diplomatic outlook: The meeting demonstrated both progress and persistent challenges in US-China relations regarding technology and security.

• Both countries emphasized the importance of international cooperation in AI development
• China explicitly warned against a “new Cold War” and rejected containment strategies
• The agreements on AI principles represent a rare point of alignment in an otherwise strained technological relationship

Looking ahead: While this initial agreement on AI governance principles marks progress, the persistent cybersecurity disputes and ongoing tensions over military technology applications suggest that comprehensive US-China cooperation on AI development remains a complex challenge that will require continued diplomatic engagement.