BIS Should Clarify Whether Chip Controls Apply to AI-Capable CPUs, Researchers Say
The Bureau of Industry and Security should clarify whether new export controls aimed at preventing China from obtaining advanced computing chips apply to artifical intelligence-capable central processing units (CPUs), researchers with Georgetown University’s Center for Security and Emerging Technology said.
Sign up for a free preview to unlock the rest of this article
Export Compliance Daily combines U.S. export control news, foreign border import regulation and policy developments into a single daily information service that reliably informs its trade professional readers about important current issues affecting their operations.
The restrictions, which were released in October 2022 and expanded a year later (see 2310170055), were intended to apply to high-performing AI chips, but they increasingly ensnare CPUs, also known as general-purpose processors, because CPU designers increasingly incorporate specialized elements for AI computation, the researchers said.
“Thus far, BIS has not clarified its policy stance on whether the latest semiconductor controls should apply to general-purpose processors,” the authors wrote. “As these processors continue to advance, BIS must grapple with the geopolitical, economic, and national security trade-offs of applying a country-wide control not just on advanced AI chips destined for China but also on an increasingly wider set of CPUs.”
As BIS weighs the issue, it should keep in mind that carving out CPUs from export controls likely would make more chips available to China for AI deployment, the researchers said. However, if a carve-out is not provided, Chinese CPU designers could gain additional revenue, potentially accelerating their development of domestic alternatives to U.S. CPUs.
“In the long term, these dynamics risk undermining U.S. leverage and could hamper U.S. CPU designers’ competitive edge,” the authors wrote. “This incentive will only grow as CPUs continue to advance and increasingly pass the control thresholds.”
The post was written by CSET data research analyst Jacob Feldgoise, research analyst Hanna Dohmen and software engineer Brian Love.