In a sharp turn from the rhetoric of outright bans, Washington has brokered a deal allowing two of its leading semiconductor firms to resume sales of customised AI chips to China ~ on the condition that 15 per cent of the resulting revenue flows directly to the US government.
These are not the flagship chips once deemed too advanced to cross the Pacific, but specially tailored, less powerful versions built to meet previous export restrictions. On paper, the arrangement offers a neat compromise: American firms reclaim access to one of their most important markets, China gains the technology it wants, and Washington collects a significant cut ~ potentially up to $2 billion ~ while keeping the most cutting-edge chips out of Beijing’s reach. Yet the deal’s novelty lies less in its economic arithmetic and more in the political precedent it sets.
The US has long restricted high-tech exports to strategic rivals in the name of national security. This approach was meant to slow potential adversaries’ technological gains, especially in fields like artificial intelligence that can blur the lines between civilian and military use. The reversal suggests that commercial interests and geopolitical bargaining have taken priority over the earlier security calculus. If the same chips were once considered a threat, what has changed other than the prospect of revenue?
For investors and industry, the uncertainty is as troubling as the restrictions themselves. Semiconductor companies plan years ahead, betting billions on product development and fabrication. Sudden political pivots ~ first a ban, then a negotiated levy ~ inject volatility into forecasts. Shareholders may welcome renewed Chinese sales, but they must also reckon with the possibility that political winds could again shift abruptly. For policymakers, the move signals willingness to monetise rather than simply restrict technology flows to rivals.
This raises larger questions: Will other sectors, from electric vehicles to consumer electronics, now face similar “market access tolls” when dealing with China? Could this become an informal tax on strategic trade, negotiated case by case? And if so, does it strengthen US leverage, or dilute the coherence of its national security stance? China, meanwhile, benefits from restored supply at a time when its AI ambitions are expanding rapidly. For Washington, the deal may also serve as a bargaining chip in unrelated areas such as rare-earth minerals, where Beijing holds a near-monopoly.
This transactional flexibility may yield short-term wins, but it risks projecting an image of policy driven by opportunism rather than principle. The chip agreement is not simply a trade story. It is a marker of how the tech rivalry between the world’s two largest economies is evolving. What began as a straightforward contest to block or protect strategic technologies is morphing into a complex negotiation where security concerns, market access, and government revenue share the same table. In this new phase, the lines between protection, profit, and policy are becoming harder to tell apart.