America’s AI ‘Sputnik moment’

With last week’s news of China’s DeepSeek seemingly overtaking the biggest American artificial intelligence companies with a rival model that is reportedly better and created at a fraction of the price, the US might have experienced its AI “Sputnik moment.” Just like the competition between the Americans and Russians during the Cold War and the success of Russia in deploying the first satellite into orbit in 1957, which fueled a space race that America won by landing on the moon, we might see a similar American drive to win this rivalry with China and preserve its position as the world’s AI leader.
DeepSeek has already created a maelstrom. It has shaken consumer confidence in the expensive American AI models, upended the markets and had investors rushing to sell their stock. But most importantly, it has raised the heat of the Chinese-American rivalry, adding to the uncertainty in the relationship that comes with a new American president. The New York Times called it the “darkest hour” in Silicon Valley, borrowing the words of British Prime Minister Winston Churchill during the Second World War.
The news alarmed the new Trump administration regarding China’s prowess as an AI competitor, despite all the efforts America has taken to limit its rival’s access to its technology, especially the semiconductors needed for the building of a successful AI model. President Donald Trump said that the rise of DeepSeek “should be a wake-up call for our industries that we need to be laser-focused on competing.”
This came as Trump was trying to handle another problem, the future of the popular Chinese video-sharing app TikTok. But Trump struck an optimistic note, while acknowledging the challenge posed by DeepSeek, when he said that he still expects US tech companies to dominate the AI industry.
DeepSeek dominated Apple’s App Store, overtaking OpenAI’s ChatGPT, and became the leading story not only in the US but around the world. It raised questions about the fundamentals of investing in America’s AI models because the Chinese model showed that the American companies’ approach and the large investments in infrastructure might have been misguided and needlessly expensive.
For example, OpenAI CEO Sam Altman was quoted as putting the cost of training its GPT-4 chatbot at $100 million. The same focus on spending big and prioritizing infrastructure was announced by Microsoft President Brad Smith last month, when he announced that the company was on “track to invest approximately $80 billion to build out AI-enabled data centers to train AI models and deploy AI and cloud-based applications around the world.”
In comparison, DeepSeek, which was established in 2023, boasted that its chatbot cost merely $5.6 million to build and train. As a result of this disparity in approach and price tag, with the billions that American AI companies spend on infrastructure, powerful computers and processors to build their large language models, DeepSeek has sent shock waves through the technology world, as well as the stock market.
The primary victim was Nvidia, which saw its stock fall almost 17 percent, losing nearly $600 billion in market value. Broadcom faced a similar fate, with a 17.4 percent decline in its stock value.

DeepSeek has shaken consumer confidence in the expensive American AI models, upending the markets.

Dr. Amal Mudallali

The bloodbath came just days after Trump, on his second day in office, announced an initiative to create a new company, Stargate, to invest $500 billion to boost AI infrastructure in the US. America’s big investors in AI, such as Microsoft and Alphabet, lost out in the market sell-off, while Oracle, one of the four participating companies in the president’s initiative, saw its shares lose 13.8 percent of their value. The other three partners are OpenAI, SoftBank and MGX.
What happened will no doubt make the US reconsider its AI strategy of spending big and focusing on massive infrastructure.
Experts have started casting doubt on whether bigger is better. The Chinese startup showed that America’s tech companies are unnecessarily making AI an exclusive club for the rich and powerful, while leaving everybody else behind. What happened, experts noted, will put pressure on US AI companies to “justify their high expenditures,” and they will be forced to lower their fees to be competitive, according to The Hill newspaper.
Astronomical spending on tech projects is not exclusive to AI companies, however. It also happens in the space sector where, for example, NASA’s VIPER program cost $433.5 million before it was scrapped in 2024, while India’s Chandrayaan-3 landing on the moon’s south pole cost only about $75 million.
But China is investing big in AI and last May it was reported that it planned to set up its largest ever state semiconductor investment fund worth $47.5 billion. The US cannot play catchup and this latest episode will put the competition between the two powers on a new trajectory, which could transform the AI race that the American side is determined to win. Washington has tried to slow China’s AI progress, but the events of last week show it might have failed to do so.
DeepSeek’s success has raised questions about the effectiveness of the American government’s export controls on China. The Biden administration announced these controls last year in an effort to limit Beijing’s access to AI technology, especially semiconductors and advanced chips.
DeepSeek presents a challenge to both the American government and to the country’s high-tech and AI companies. If DeepSeek’s success proves to be serious and lasting — though Elon Musk is skeptical and has cast doubt on this success — the American companies will have to change their strategy. The Chinese company has proved bigger is not always better and, as SalesForce CEO Marc Benioff said, “AI’s true value lies in data, not expensive hardware.”
President Trump has to refocus America’s AI trajectory and let these companies learn from Musk as to how it can be done cheaper. That way, Americans will not break the bank trying to win the AI race, while China breaks new records in making the technology cheaper and maybe even better.
But the biggest takeaway from the DeepSeek news is that it has set the democratization of AI on a new path that others will not be able to ignore or reverse. AI models and technology will become more accessible to small startups and ordinary people alike, even if they have little or no money.
As for the US and its competition with China, it is important to remember from the space race that Sputnik was followed by Russia making Yuri Gagarin the first human in space. Trump’s challenge now is to have his own Gagarin, otherwise China will dominate AI globally to America’s detriment.

• Dr. Amal Mudallali is an international affairs adviser for Think and a former Lebanese ambassador to the UN.