What's happening
The US Commerce Department on May 31, 2026, issued formal guidance establishing that export license requirements apply to advanced AI chips destined for entities headquartered in China, regardless of whether those entities are physically located outside Chinese territory. The directive specifically implicates Nvidia's Rubin and Blackwell GPU architectures and AMD's MI350x AI accelerators ��� among the most capable commercially available chips for large-scale AI workloads. The guidance was reported by Reuters on May 31, 2026.
The enforcement action addresses a gap that opened in May 2025, when the Commerce Department opted not to enforce the AI Diffusion rule. During the intervening period, industry sources estimate that hundreds of thousands of chips may have flowed to Chinese subsidiaries operating abroad through that gap. Chris McGuire, a technology expert and former State Department official, characterized the scale of the problem in stark terms: 'This is a HUGE problem. Chinese companies have been buying these chips, very likely at scale.'
Why it matters for markets
Nvidia, with annual revenue of $253.49 billion and a market capitalization of $5.11 trillion, derives substantial business from data center customers globally, and its Blackwell GPU family has been central to AI infrastructure buildouts. Any licensing regime that restricts Chinese-headquartered entities — including their overseas subsidiaries — from accessing Nvidia hardware without prior government approval introduces friction into a customer segment that has historically represented meaningful demand. AMD, with annual revenue of $37.45 billion and a market cap of $841.55 billion, faces analogous constraints on its MI350x Instinct accelerator line, which competes directly in the AI and high-performance computing market.
The breadth of the new guidance — extending jurisdiction to overseas subsidiaries of Chinese-headquartered firms — is notable because it closes a structural workaround that had allowed procurement to continue even as direct China-bound shipments faced tighter scrutiny. The estimate of hundreds of thousands of chips potentially having moved through this channel during the enforcement gap suggests the commercial volumes at stake are not marginal. For both Nvidia and AMD, the practical effect is that a wider universe of prospective customers now requires individual export licenses, a process that adds regulatory uncertainty to sales cycles.
The new licensing requirement does not constitute an outright ban, but the addition of a government approval layer for transactions involving Chinese-headquartered entities — wherever they are domiciled — materially changes the compliance calculus for both chip designers and their distribution partners. Companies seeking to purchase Nvidia or AMD AI accelerators will need to demonstrate they fall outside the scope of the guidance or obtain the requisite licenses, potentially slowing procurement timelines.
Sectors and assets to watch
The primary tickers directly affected are NVDA (Nvidia Corporation, market cap $5.11 trillion) and AMD (Advanced Micro Devices, market cap $841.55 billion), both of which design the specific chip families named in the Commerce Department guidance — Nvidia's Rubin and Blackwell GPUs and AMD's MI350x accelerators. Both companies trade on the Nasdaq and are the dominant US-based designers of AI accelerator hardware, making them the most immediate focal points for any market or operational response to the new export enforcement posture.
Beyond the two named chip designers, the guidance has implications for the broader AI hardware supply chain, including cloud infrastructure providers, data center operators, and semiconductor distributors that serve Chinese-headquartered clients with international operations. Companies in the AI diffusion ecosystem — whether as resellers, system integrators, or hyperscale customers — will need to assess whether their end customers fall within the newly enforced definition of restricted entities.
What to watch next
Key developments to monitor include whether the Commerce Department issues further clarifying guidance on the precise definition of 'headquartered in China' and how it applies to complex multinational corporate structures, as well as whether Nvidia or AMD make formal disclosures quantifying the revenue exposure associated with the newly restricted customer segment. Investors and analysts will also be watching for any diplomatic response from Beijing, potential retaliatory measures affecting US technology companies operating in China, and whether the license application process results in approvals or denials for specific transactions — outcomes that would signal how restrictive enforcement will be in practice.