The AI Competition: Democratic Innovation vs. Directed Development in US-China Strategic Rivalry

The competition for AI leadership represents one of the defining technological contests of the 21st century, with consequences that will ripple across military modernization, industrial strategy, and global influence. China's approach—characterized by state oversight, coordinated public-private development through firms like Baidu, Alibaba, Tencent, and DeepSeek, and systematic integration of civilian and defense research—presents both a competitive challenge and a fundamentally different model for technological advancement. This discussion assessed the strengths and vulnerabilities of competing approaches, examined how different data governance regimes and development methodologies affect trusted deployment across civilian and military applications, and explored policy frameworks for maintaining democratic technological leadership while coordinating responses with allies facing similar strategic challenges in the AI domain.

This event featured John Costello, an intelligence analyst and former senior official across both the Trump and Biden Administrations, including as Deputy Assistant Secretary of Commerce for Intelligence and Security and the inaugural Chief of Staff of the Office of the National Cyber Director; and Dr. Thomas Șerban von Davier, an artificial intelligence and machine learning research scientist at Carnegie Mellon's Software Engineering Institute who researches frontier AI, open-source large language models, and international AI development trends. Ralph López, CMIST Deputy Director for Security Policy Studies, moderated the discussion.



About our Speakers:

Thomas Şerban von Davier is an AI research scientist at the Carnegie Mellon University Software Engineering Institute (SEI) and a CMIST faculty affiliate. He is a member of the Frontier Lab researching the cutting edge of AI engineering and interaction. He has collaborated with the Oxford Centre for Technology and Development as a specialist on artificial intelligence. Before that he worked in the technology sector as a data engineer. Thomas holds a DPhil in Computer Science from the University of Oxford, a Masters in Human-Computer Interaction from Carnegie Mellon University, and a BS in Psychology and Human-Computer Interaction also from Carnegie Mellon University.

John Costello is a non-Resident Fellow at Carnegie Mellon Institute for Strategy & Technology, He previously served as the first chief of staff and principal architect of the Office of the National Cyber Director (ONCD), the first White House component created since 1988. Prior to his time at ONCD, Costello served as the deputy executive director of the Cyberspace Solarium Commission, where he led the commission’s work on cybersecurity, resilience, and supply chain security, as well as the commission's legislative efforts. Costello served as the deputy assistant secretary of intelligence and security at the Department of Commerce. In that capacity, Costello established the Commerce Department’s intelligence analysis capability and led the department’s information and communications technology and service supply chain analytical and regulatory efforts. Costello is a US Navy veteran, serving his tours of duty at the National Security Agency as an intelligence analyst and operations officer. While at the National Security Agency, he specialized in foreign adversary cyber, electronic, and information warfare. John has a working professional proficiency in Mandarin Chinese and graduated with honors from the Defense Language Institute.




Originally taped on 18 November 2025 at Carnegie Mellon University as part of CMIST's Scientists and Strategists Speaker Series

Video © Carnegie Mellon Institute for Strategy and Technology
Presenter Materials © John Costello and ©Thomas Şerban von Davier


*All content herein is the intellectual property of Carnegie Mellon Institute of Strategy and Technology (CMIST). Unauthorized downloading, reproduction, rebroadcasting, or distribution of any video content is strictly prohibited. These videos are provided for educational purposes only. Any commercial use or further distribution without express written permission from CMIST is strictly forbidden. By accessing these videos, you agree to the terms and conditions of use, which prohibit any copyright infringement or misuse of the content.