China is embracing “embodied AI”—artificial intelligence integrated with physical agents, such as robots and drones—both for commercial reasons and as a path to artificial general intelligence (AGI). The trend reflects China’s signature approach to AI, which recognizes diverse paths to AI dominance vis-à-vis the large models favored in the United States. This report documents PRC support for AI embodiment, describes how it is understood by China’s research community, and maps out the related infrastructure.

System-to-model innovation is an emerging innovation pathway in artificial intelligence that has driven progress in several prominent areas over the last decade. System-level innovations advance with the diffusion of AI and expand the base of contributors to leading-edge progress in the field. Countries that can identify and harness system-level innovations faster and more comprehensively will gain crucial economic and military advantages over competitors. This paper analyzes the benefits of system-to-model innovation and suggests a three-part framework to navigate the policy implications: protect, diffuse, and anticipate.

Wuhan, China’s inland metropolis, is paving the way for a nationwide rollout of “embodied” artificial intelligence meant to fast-track scientific discovery, optimize production, streamline commerce, and facilitate state supervision of social activities. Grounded in real-world data, the AI grows smarter, offering a pathway to artificial “general” intelligence that will reinforce state ideology and boost economic goals. This report documents the genesis of Wuhan’s AGI initiative and its multifaceted deployment.

Maintaining long-term U.S. leadership in artificial intelligence will require policymakers to foster a diversified, contestable, and competitive market for AI systems. Today, however, incumbent technology companies maintain a distinct advantage in the production of large AI models, and they have the means and motion to use their control over key chokepoints in the AI supply chain (compute, data, foundation models, distribution channels) to stifle competition. This report explores the associated economic and national security risks, and offers recommendations for maintaining an open and competitive AI industry.

Helen Toner testified before the House Judiciary Subcommittee on Courts, Intellectual Property, Artificial Intelligence, and the Internet on recommendations to bolster security and transparency around U.S.-developed frontier AI.

We investigate the scale of attack and defense mathematically in the context of AI's possible effect on cybersecurity. For a given target today, highly scaled cyber attacks such as from worms or botnets typically all fail or all succeed.

In response to the Office of Science and Technology Policy's request for input on an AI Action Plan, CSET provides key recommendations for advancing AI research, ensuring U.S. competitiveness, and maximizing benefits while mitigating risks. Our response highlights policies to strengthen the AI workforce, secure technology from illicit transfers, and foster an open and competitive AI ecosystem.

Large generative models are widely viewed as the most promising path to general (human-level) artificial intelligence and attract investment in the billions of dollars. The present enthusiasm notwithstanding, a chorus of ranking Chinese scientists regard this singular approach to AGI as ill-advised. This report documents these critiques in China’s research, public statements, and government planning, while pointing to additional, pragmatic reasons for China’s pursuit of a diversified research portfolio.

Sam Bresnick testified before the Senate Judiciary Subcommittee on Privacy, Technology, and the Law regarding tech companies' ties to China and their implications in a future conflict scenario.