CUMNOCK, Scotland — King Charles III opened the day by asking some of the artificial-intelligence industry’s most powerful figures whether it was “all too late” to preserve human control over the technology they were building.
Hours later, Jensen Huang stepped away from the dignitaries gathered at Dumfries House and told reporters that Nvidia expected to sell twice as many chips the following year.
The contrast on September 17 was difficult to miss. The king’s summit at his historic Ayrshire estate, convened by the King’s Foundation and the Sustainable Markets Initiative, was intended to challenge AI leaders on safety and accountability.
Huang arrived with both a warning and one of the most bullish forecasts Nvidia has issued.
“I expect Nvidia to sell twice as many chips this next year as we do this year,” he said on the sidelines of the gathering, which drew OpenAI chief financial officer Sarah Friar, Google DeepMind co-founder and chair Demis Hassabis, and Anthropic’s public affairs chief Tino Cuellar. The last of those names underlined an awkward irony: Anthropic, the AI safety company that argues the technology poses existential risks, this week disclosed plans for a potential $2 trillion IPO that would make it one of the world’s most valuable companies.
The forecast Huang issued marks a significant upward shift from Nvidia’s August earnings guidance. At its most recent quarterly release, the company projected revenue growth of approximately 70 percent for the next fiscal year. The fine print mattered: Nvidia disclosed at the same time that revenue could double, contingent on supply meeting actual demand. Huang appeared to close that gap on Wednesday. “We’ve never forecasted like this before,” he said. “We have a lot greater visibility now.”
“Actual demand from the AI industry is much greater than 70 percent,” he added, describing the sector as having “reached its inflection point.” The demand, he explained, was not coming from one geography or one industry: “AI has so much contribution to the benefits of different industries, different economies, and you can see that in almost every single country that we’re in, people want to invest in AI.”
Huang was careful to maintain at least a rhetorical balance with the summit’s stated purpose. He framed safety as fundamentally an engineering challenge. “When a product is not safe, we should hold it back and keep engineering it,” he told attendees, a formulation that places the AI industry in charge of its own progress. King Charles, for his part, struck a different register in his opening remarks, telling the executives: “Those in our world who value our humanity and its vital moral component are anxiously seeking your reassurance that we will not lose control of our destiny.”
The substance of Huang’s announcement will matter most to the countries and institutions scrambling to secure their place in the AI supply chain. As Seeking Alpha reported this week, Nvidia’s updated guidance signals a production ramp on Blackwell chips that significantly exceeds what the company was willing to forecast as recently as August. For nations racing to build sovereign AI infrastructure across the Gulf, South Asia, and Southeast Asia, the question was never whether demand existed. It was whether supply would follow.
Nvidia’s May earnings release offered a window into how far the demand had already traveled. The company posted $81.6 billion in first-quarter fiscal 2027 revenue, an 85 percent jump from the prior year, driven almost entirely by data center orders. Huang used that report to describe the AI buildout as “the largest infrastructure expansion in human history.” His Scotland remarks suggest the next year will exceed even that framing.
The distribution of that doubling, however, is not uniform. US government restrictions on advanced AI chip exports to China remain in force, and Nvidia has said it assumes no significant data center compute revenue from China in the current quarter. The trade standoff, which Fortune traced to the broader contest between Washington and Beijing over AI primacy, has redirected chip supply into markets outside the export-control perimeter, giving countries not subject to US export controls a structural advantage in securing Blackwell allocations.
Where those chips land, and who controls the AI systems that run on them, is a question that the broader US-China technology standoff made significantly more complicated earlier this year. The competition extended into cloud access and AI model licensing, reshaping the global hardware market along geopolitical lines that predate the current chip cycle.
NVDA stock rose roughly two percent in the session following Huang’s Scotland remarks, a measured response that suggested investors read the doubling as credible, not surprising. The August earnings language had prepared the market; Wednesday removed the conditional.
What it did not remove was King Charles’s underlying question. Doubling chip production in twelve months does not resolve who governs the systems those chips enable, or whether the technology’s pace can be matched by the institutional capacity to understand it. The king’s summit will produce a communiqué. Huang’s forecast will produce a product cycle. Which of the two has a more lasting effect on the AI future he described may not be answerable before the next generation of hardware ships.
