The recent announcement of a bilateral channel for AI consultation between President Joe Biden and General Secretary Xi Jinping opens up new possibilities for navigating the complex landscape of artificial intelligence. As the U.S. and China embark on this dialogue, focusing on specific, achievable issues is crucial to making meaningful progress. This article outlines a roadmap for the U.S.-China AI dialogue, emphasizing key areas of discussion and cooperation.
Photo from: Bloomberg.com
Three Key Focus Areas
When President Biden and General Secretary Xi met in November 2023, they highlighted the need for a collaborative approach to address the risks and enhance the safety of advanced AI systems. The challenge lies in defining the specific issues at hand, given the broad spectrum of AI applications, ranging from self-driving cars to potential military applications. This article proposes a focused approach to ensure the dialogue yields tangible results.
In order to ensure the effectiveness of the U.S.-China AI dialogue, Graham Webster and Ryan Hass suggest a focused approach with prioritization in three key areas. The first area involves addressing the impact of AI on U.S.-China strategic stability, establishing boundaries and expectations for acceptable military uses, collaborating on rigorous testing and evaluation standards for AI-enabled military systems, and pursuing early wins such as mutual agreement that nuclear launch decisions should never be automated.
The second key area emphasizes the importance of enabling necessary data sharing. This involves recognizing the economic value and security vulnerability associated with large datasets, exploring the use of privacy-enhancing technologies to facilitate secure data sharing, and developing standards specifically tailored to data sharing in sectors like climate research or disease control.
Lastly, the authors highlight the significance of addressing common challenges and establishing trusted standards. This includes collaboration on labeling or watermarking outputs of generative AI models, exploring both technical and policy solutions for maintaining information integrity, and actively involving industry and global standards organizations in the development and implementation of common standards. This structured approach aims to streamline the dialogue and promote practical solutions to the complex landscape of AI collaboration and governance.
READ ALSO: US And Chinese Military Resume Talks In Effort To Mitigate Growing Tensions
Staying Focused and Avoiding Pitfalls
The authors caution against swerving into intractable issues that might derail the dialogue’s potential. While acknowledging concerns such as export controls and domestic surveillance, the roadmap suggests keeping the focus on issues that offer practical governance and risk reduction solutions. Avoiding zero-sum discussions and emphasizing early wins will be essential to maintaining momentum and ensuring the value of the dialogue for both nations.
As the U.S. and China navigate the uncharted impact of machine-learning models, the proposed roadmap offers a structured approach to advancing the dialogue. By concentrating on specific, achievable objectives, both nations can seize the opportunity to foster cooperation, reduce risks, and shape the governance of emerging technologies in a rapidly evolving landscape.
