SAFE Chips Act: New Export Ban Automatically Denies Licenses for Advanced Chips to Foreign Adversaries

The new Secure and Feasible Exports of Chips Act of 2025—the SAFE Chips Act—is about drawing a hard line on who gets access to America’s most advanced computing power. Simply put, this bill creates a mandatory export ban on high-performance semiconductors, primarily aimed at preventing them from reaching designated foreign adversary countries. For anyone working in tech, manufacturing, or international trade, this is a major regulatory shift that replaces licensing review with automatic denial.

The Automatic Stop Sign: What Gets Banned

Starting immediately upon enactment, the Secretary of Commerce must require a license for exporting, reexporting, or transferring any “advanced integrated circuit” to a foreign adversary country or to any entity whose parent company is headquartered there (SEC. 2). Crucially, the bill mandates that any application for such a license must be denied. This isn't a case-by-case review; it’s an automatic stop sign. For U.S. chip makers and exporters, this closes off significant markets overnight and removes any discretion the Commerce Department previously had in approving these specific sales.

Defining “Advanced”: The Technical Specs

What exactly is an “advanced integrated circuit”? The bill defines it using highly technical parameters, essentially targeting the chips used for high-end AI and supercomputing. It includes any product classified under specific Export Control Classification Numbers (ECCNs), like 3A090 or 4A090, or anything “functionally equivalent or substantially similar.” If you're not in the semiconductor industry, think of it this way: if a chip meets or exceeds performance thresholds like a total processing performance of 4,800 or a combined bandwidth of 5,000 gigabytes per second (SEC. 2), it’s banned. These are the chips that power major data centers and advanced military or cyber capabilities.

There is one key exception: the ban does not apply to chips that are not designed or marketed for use in data centers. This is where things get a little squishy. If a high-performance chip can be used for multiple purposes (a dual-use technology), its export status might hinge on how the company markets it, rather than its technical capability. This subjective element could create enforcement gray areas for both companies and regulators.

The Real-World Impact on Business

For U.S. technology companies, this means a significant chunk of their potential international market is off-limits. If you manufacture these high-end chips, you lose the ability to sell to customers in countries like China, Hong Kong, and Macau, as they are included in the definition of a “foreign adversary country.” This could force companies to rapidly re-evaluate their sales forecasts and supply chains, potentially leading to a surplus of specialized components or a pivot to less restricted markets.

On the flip side, the bill grants the Secretary of Commerce the power to update these technical parameters after 30 months. Before any change, the Secretary must brief Congress on how the update would affect the capabilities of leading Chinese AI firms and how the U.S. will maintain its “computing advantage” (SEC. 2). This ensures that the definition of “advanced” can keep pace with technology, but it also means the regulatory goalposts could shift, adding an element of long-term uncertainty for manufacturers planning future product lines.