New CNAS Research Proposes Solutions to Secure AI Chip Supply Chain

Washington, December 11, 2024 – Today, the Center for a New American Security (CNAS) released Technology to Secure the AI Chip Supply Chain: A Working Paper, by Tim Fist, Tao Burga, and Vivek Chilukuri.

The paper explores the challenges of securing the AI chip supply chain and proposes innovative hardware-enabled mechanisms (HEMs) to balance U.S. economic competitiveness, national security, and democratic values.

The report highlights that existing export controls on AI chips are difficult to enforce and risk becoming overly burdensome on exporters and end users. These controls are also vulnerable to evasion—through shell companies—creating significant gaps in enforcement.

The authors propose leveraging HEMs—security mechanisms within data center AI hardware—as a more targeted solution to address these challenges. These mechanisms are already widely used in commercial and defense products, such as Apple’s Secure Enclave and Google’s chip verification systems and offer a promising pathway to enhance AI governance and security.

To advance the development and implementation of HEMs, the report recommends that U.S. policymakers:

• Accelerate HEM and hardware security research and development (R&D) through direct funding and public-private partnerships,

• Create commercial incentives for industry HEM R&D through conditional export licensing, and

• Further develop AI hardware security standards to incentivize and harmonize security features across industry.

The report underscores the need for innovative approaches to secure the AI chip supply chain without stifling innovation or imposing undue surveillance burdens. It provides a roadmap for policymakers to address these challenges while fostering a robust and secure technological landscape.

For more information or to arrange an interview with the report’s authors, please contact Alexa Whaley at [email protected].