On January 13, 2025, the U.S. government announced new regulations that significantly tighten controls on the export of artificial intelligence (AI) chips and related technology.
The updated rules aim to maintain the U.S.'s leading position in AI while preventing adversaries, including China, Russia, Iran, and North Korea, from accessing advanced computing power.
Countries are categorized into tiers, restricting AI chip exports to most nations, while allowing unrestricted access to close allies such as Japan, the United Kingdom, and South Korea.
U.S. Commerce Secretary Gina Raimondo stated, "The U.S. leads AI now -- both AI development and AI chip design, and it's critical that we keep it that way." This regulation concludes a four-year effort by the Biden administration to limit China's development of advanced AI technologies.
The measures include restrictions on the export of advanced graphics processing units (GPUs), essential for training AI models. Companies like Nvidia and AMD are expected to be significantly impacted, with their shares falling by 2% to 3% in premarket trading.
Major cloud service providers, including Microsoft, Google, and Amazon, also experienced slight stock price declines following the announcement.
U.S.-headquartered companies such as AWS and Microsoft will be allowed to deploy no more than 50% of their AI computing power outside the U.S., with stricter limits for Tier 2 and Tier 3 countries.
While the Biden administration defends the regulations as necessary for national security, industry leaders express concerns about potential negative impacts on business. Nvidia has criticized the measures as a "sweeping overreach" that could hinder innovation.
Data center provider Oracle warned that these rules might inadvertently strengthen Chinese competitors by restricting access to global markets.
The regulations also introduce licensing requirements for certain chips and control the export of "model weights," critical components in AI models.
National Security Adviser Jake Sullivan emphasized the need for the U.S. to prepare for rapid advancements in AI capabilities, which could significantly affect the economy and national security.
The regulations are set to take effect 120 days after the announcement, allowing for input from the incoming administration under President-elect Donald Trump.