The Trump administration is considering worldwide export controls on AI chips — restrictions that would directly hit Nvidia and AMD, the two U.S. designers whose accelerators dominate global AI training and inference hardware.

According to a report from The Motley Fool, the prospective measures would extend American export jurisdiction beyond specific countries of concern and apply to AI chip shipments on a worldwide basis. For Nvidia and AMD, that framing matters: it converts export licensing from a country-by-country question into a global one, with Washington as the gatekeeper for virtually every advanced accelerator sale.

What would the controls change?

Under the existing regime, Nvidia and AMD already operate under layers of restriction. Advanced accelerators face licensing requirements for shipments to China and other designated destinations, and successive U.S. rulemakings have tightened the performance thresholds that determine which products require licenses. The companies have responded by re-engineering products to sit just below control thresholds for markets they can still legally serve.

A worldwide control regime would remove that workaround. If every AI chip export — regardless of destination — requires a U.S. license, Nvidia and AMD lose the ability to redirect restricted silicon to unrestricted markets. Data center customers in countries with no current restrictions on American AI hardware would suddenly sit behind the same licensing wall as restricted destinations.

The Motley Fool's report frames this as a possibility under consideration by the administration, not a finalized rule. Companies, customers, and investors are left pricing in a regulatory outcome that has no announced effective date, product scope, or licensing standard.

Why the stakes concentrate in two companies

Nvidia and AMD sit at the center of this policy question because their GPU families power the overwhelming majority of large-scale AI compute deployed globally. Any worldwide control regime would, in practice, function as an Nvidia-and-AMD licensing system — with knock-on effects for the hyperscalers, cloud providers, and sovereign AI programs that buy their hardware.

For both companies, the exposure runs through the data center segment, which now anchors their revenue. Restrictions that slow or condition accelerator shipments to any destination translate directly into demand uncertainty for their highest-value product lines.

The reported deliberation also lands amid a broader shift in U.S. export policy posture, where AI chips have become the primary instrument of technology competition. Each previous tightening of rules has forced product-line adjustments and re-segmentation of the market into compliant and non-compliant tiers.

What to watch next

The key variables that will determine the commercial impact are the same ones undefined in the report: which products fall under a worldwide control, which countries receive license exceptions if any, and how quickly the administration moves from consideration to rulemaking. Nvidia and AMD have navigated successive rounds of restrictions by redesigning products and re-routing sales; a worldwide regime would close off the second option entirely. Until the administration publishes an actual rule, the market is trading on the possibility that the next escalation in AI chip policy removes geography as an escape valve.