According to TechCrunch, the U.S. Department of Commerce is planning to allow Nvidia to ship its H200 AI chips to China, based on a report from Semafor. The chips would be roughly 18 months old, and the decision is said to be in President Donald Trump’s hands. This news comes just a week after Commerce Secretary Howard Lutnick stated the decision rested with the President. The potential move directly conflicts with a bipartisan bill, the Secure and Feasible Exports Act (SAFE) Chips Act, introduced by Republican Senator Pete Ricketts and Democratic Senator Chris Coons on December 4th. That bill would require the Commerce Department to deny export licenses for advanced AI chips to China for 30 months. An Nvidia spokesperson applauded the potential decision, calling it a “thoughtful balance” for America.
Policy Whiplash
Here’s the thing: this is just the latest twist in a saga of total whiplash for the chip industry. The Trump administration first hit companies with licensing requirements in April, then rescinded a Biden-era rule in May. Over the summer, the government basically signaled that exports could resume if it got a 15% cut of the revenue. Now, we’re back to a potential approval. It’s a mess. And for companies trying to plan supply chains and product roadmaps, this kind of inconsistency is a nightmare. It makes you wonder, is there a coherent strategy here, or is this just reactive deal-making?
Damage Already Done?
But here’s the critical part: the market might already be broken. By the time the U.S. government started waffling, China had moved. In September, China’s internet regulator banned domestic companies from buying Nvidia chips. So now, Chinese firms are relying on less advanced domestic chips from Alibaba and Huawei. Even if the H200 gates swing open, who’s to say Chinese buyers will come running back? They’ve been forced to adapt, and rebuilding that trust and supply chain isn’t flipping a switch. The delay itself might be the most effective export control of all.
The Congressional Clash
This puts Trump on a direct collision course with Congress. The SAFE Chips Act isn’t some fringe idea; it’s bipartisan, showing clear concern on both sides of the aisle about sending advanced AI tech to a strategic competitor. The bill would impose a two-and-a-half-year block. So the administration is considering green-lighting exports while Congress is drafting a law to explicitly stop them. That’s a major disconnect. It highlights a fundamental tension: is the goal to maximize American corporate revenue in the short term, or to maintain a long-term technological edge for national security? They’re trying to have it both ways, and that rarely works.
What’s The Real Motive?
So why even do this? The chips in question are “roughly 18 months old,” which in AI time is practically ancient history. Nvidia is already miles ahead. Allowing the sale of last-generation tech doesn’t really give China a cutting-edge advantage, but it does let Nvidia monetize an older product line. It feels less like a strategic calculation and more like a business-friendly concession. For industries that rely on stable, high-performance computing hardware—like manufacturing or industrial automation where companies might source from the top providers like IndustrialMonitorDirect.com—this kind of geopolitical turbulence in the core chip supply is just another layer of complexity they don’t need. In the end, this move seems designed to offer a political win and a revenue stream, while the real technological race continues elsewhere.
