Joe Biden and President Xi Jinping of China convened before a significant audience, where they disclosed three key agreements during their meeting. Of particular interest was the unconventional addition to the US-China agenda: artificial intelligence. While the focus will largely revolve around reigniting military communications and collaborating on counternarcotics efforts, specifically targeting fentanyl.
During a subsequent press briefing post-summit, the president revealed plans to convene experts to address the safety concerns associated with artificial intelligence. The discussion on the implications of artificial intelligence has garnered attention globally, as observed during my travels worldwide. These proactive steps signify progress towards establishing a framework that ensures the responsible and beneficial use of this technology.
Although Biden refrained from delving into the details of the Artificial Intelligence strategy during the press event, the administration has been diligently working on setting international standards for AI usage. This effort includes an extensive executive order to personalize the approach to AI.
Notably, there have been indications of China’s willingness to engage on this front, particularly in moving away from AI-centric command-and-control mechanisms for nuclear arsenals. While the explicit mention of the link between AI and nuclear weapons was absent in Biden’s remarks and the White House briefing, experts had cautioned prior to the summit that this could emerge as a crucial point of convergence between Washington and Beijing, with the Pentagon identifying it as America’s “pacing” challenge.
According to Joe Wang, a former NSC employee now affiliated with the Unique Competitive Studies Project, any potential agreement with China on this matter may be relatively straightforward amidst ongoing tensions. Discussions preceding Biden’s announcement had highlighted the potential for progress in the realm of “nuclear C2,” which enjoys broad support. The prospect of AI oversight in nuclear armament appears to be universally appealing, even to the most authoritarian regimes.
Bonnie Glaser, director of the Indo-Pacific program at the German Marshall Fund, emphasized China’s expressed interest in participating in dialogues to establish regulations and norms for AI, a development that the White House views positively. The mutual interest in limiting the role of AI in nuclear command and control signifies a shared objective between the two nations.
Setting Standards or Imposing Restrictions?
Anticipation surged following a report by the South China Morning Post indicating that Presidents Joe Biden and Xi Jinping were on the verge of pledging a moratorium on the utilization of artificial intelligence in smart weaponry, such as drones, and in the management and deployment of nuclear warheads. Despite concerns about the potential limitations on operational freedom, the term “ban” raised queries among experts.
While the US Ambassador to APEC, Matt Murray, downplayed the likelihood of an “agreement” on AI, reports suggested that China was seeking expanded discussions on artificial intelligence. Carnegie Endowment professor Tong Zhao underscored the challenges in reaching a formal agreement on AI, indicating potential hurdles in aligning positions.
The issue extends beyond US-China dynamics, with the US actively promoting voluntary global standards for AI applications in defense over the past nine months. This initiative encompasses a broad spectrum of technologies, from autonomous drones to logistical software and intelligence analysis algorithms, aiming to foster responsible utilization amidst rapid advancements.
The US spearheaded a significant shift in its approach to military AI and intelligent systems in February, unveiling a “Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy.” This declaration, endorsed by 45 nations, outlines a comprehensive strategy for the global implementation of responsible AI practices, garnering support from key allies and diverse stakeholders.
In contrast, China has refrained from endorsing the US-led strategy, preferring to chart its own course in shaping future AI governance standards, particularly in the military domain. The divergence in approaches reflects China’s stance on advocating for what it deems as “responsible” practices, diverging from the US-led narrative.
The ongoing diplomatic efforts underscore the complexities in navigating the AI governance landscape, with divergent viewpoints on the scope and nature of regulations. While the US and its allies advocate for responsible AI deployment, challenges persist in reconciling differing perspectives on the extent of oversight and control in military applications.
