According to The Wall Street Journal, Nvidia and AMD are likely to accept proposals to provide the American government with 15 percent of the revenue generated by their AI chips in exchange for continuing to sell export licenses to China. If it does happen, it would mark a significant shift in Washington’s approach to regulating the flow of advanced technologies to one of its major competitors. The strategy does not seem to impose direct bans but instead monetizes controlled exports, which may be used to finance their domestic technological base but keep their firms active in the market.
U.S. Export Control Strategy Evolves from Prohibition to Profit-Linked Access
U.S. officials historically used boundaries on exporting high-performance AI chips to bar China from accessing technology that could boost military and strategic potential. The Biden administration announced new limitations on semiconductor exports in 2022 and 2023, and the Commerce Department followed with announcements in January 2025. Such actions aimed to prevent China from obtaining advanced GPU technology.
But around the middle of 2025, policy started to mellow down. There was news that the Trump government also allowed the sale of China-specific chips, such as the Nvidia RTX PRO, that were created so as to comply with export controls but still sell to China. The purported 15 percent revenue-sharing framework is part of this trend. Instead of complete bans on items, there is a monetary scheme, which might potentially direct revenue into R&D programs in the U.S.
According to the proponents, the hybrid solution will ensure leverage over the supply chains but enable the American companies to continue enjoying profitable international business. Opponents respond that this would cause technology transfers to strategic competitors to become acceptable and thus defeat the security justification of the limits.
Global Reactions and Potential Risks in a High-Stakes Semiconductor Market
If they adopt it, revenue-sharing will be a first in the history of tech trade policy, yet it will likely attract a backlash as well. They may view cutting the tariffs by 15 percent as a form of coercion in China, which can retaliate through its tariffs or curbs on specific Chinese companies, or through an intensive internal build-up on chipmaking capacity. This might, in the long term, offset the U.S. hegemony over AI hardware as it decreases dependence on American vendors.
The planning also evokes messy legal and diplomatic issues. Traveling norms of international trade hardly allow governments to directly extract a specific proportion of private export revenues, and the World Trade Organization may consider them as discriminatory. Other countries may follow suit, and as a result, the world market will become fragmented, where entry is not based on open competition but rather political agreements.
Economically, the model provides the U.S. a manner that allows it to monetize limited exports but at the same time maintains an eye on the strategic flow of technologies. According to analysts, this is a sign of a change in economic statecraft, which is transforming critical technology into a controlled stream of revenue instead of a totally constrained resource.
Awaiting Clarity on a Potentially Precedent-Setting Policy Shift
By the 11th of August 2025, no one has conclusively established that Nvidia, AMD, and the U.S. government have a set deal. Should they establish such a deal, it might create a precedent on the balance of economic opportunity and national security in trade in the AI era. This action would represent a middle ground between an outright ban and free trade in U.S.-China technology exchange. The government will largely depend on how other nations in the global arena will view such practices and how successful they will be economically.
About The Author
