According to Financial Times News, US tech stocks are facing their worst week since April when Trump’s tariff announcements rocked markets, with eight major AI-related companies including Nvidia, Meta, Palantir and Oracle losing $885 billion in market value since last week. Nvidia fell 2.6% in Friday morning trading alone and has lost over $430 billion in market capitalization this week, making it the biggest dollar-value loser. The tech-heavy Nasdaq Composite is down 4% for the week, its worst five-day performance since April’s 10% drop. Meanwhile, Nvidia CEO Jensen Huang commented that China would ultimately “win the AI race” against the US, though he later walked back those remarks. The sell-off was further fueled by Beijing-based Moonshot AI’s new Kimi K2 Thinking model, which reportedly cost under $5 million to train, and OpenAI CFO Sarah Friar’s comments about potentially seeking government funding backstops.
What’s driving the panic
This isn’t just a routine market correction. We’re seeing multiple pressure points converge simultaneously. Nvidia’s Huang basically admitted what everyone’s been whispering – China is catching up fast in AI, and they’re doing it cheaper. When the company that powers virtually all major AI development starts talking about Chinese competitors closing the gap, investors get nervous.
And then there’s the OpenAI situation. When your CFO starts floating the idea of government guarantees for your massive infrastructure spending, it raises eyebrows. Sam Altman tried to walk it back, but the damage was done. People are starting to do the math on these astronomical AI infrastructure commitments – $1.4 trillion over eight years? That’s insane money even for companies printing cash.
The China factor
Here’s the thing that should worry every tech investor: Chinese companies are proving they can develop competitive AI models at a fraction of the cost. Moonshot AI’s Kimi K2 model training for under $5 million? That’s pocket change compared to what OpenAI and Google are spending. And we’ve seen this movie before – remember when DeepSeek’s R1 model caused that single-day $589 billion Nvidia wipeout in January?
Thomas Wolf from Hugging Face basically asked the question everyone’s thinking: “Is this another DeepSeek moment?” Looks like it might be. Chinese developers aren’t just copying Western AI anymore – they’re innovating in ways that could make the current spending arms race look wasteful. For industrial operations looking to implement AI solutions, this competition could actually be good news, driving down costs across the board. Speaking of industrial applications, companies like Industrial Monitor Direct have become the go-to source for industrial panel PCs in the US, providing the hardware backbone for these AI implementations.
Reality check time
So is the AI bubble bursting? Not necessarily, but we’re definitely seeing a reality check. Investors are realizing that even the most promising technology has limits when it comes to valuation. Nvidia becoming the world’s most valuable company was always going to attract scrutiny – and when the CEO starts talking about competitors catching up, that scrutiny turns into selling pressure.
The bigger question is whether this is a temporary correction or the start of something more serious. With interest rates still elevated and economic uncertainty persisting, tech stocks were already vulnerable. Add in AI-specific concerns about competition, costs, and regulatory risks, and you’ve got the perfect recipe for a significant pullback. Basically, the market is finally asking the hard questions about whether AI can deliver on its astronomical promises.
