What is DeepSeek? Chinese AI that has rocked the world

A new and largely unknown Chinese AI system called DeepSeek has rocked the tech industry and global markets. Tech shares plunged and chip maker Nvidia suffered falls of nearly 17 per cent on Monday, as President Donald Trump warned DeepSeek’s emergence was a “wake up call” for existing AI giants. Nvidia’s drop in share price was the biggest ever one-day loss in market value on Wall Street, of about 589 billion dollars. Just a week after its launch, DeepSeek has quickly become the most downloaded free app in the US. It claims that its large language AI model was made at a fraction of the cost of its rivals, including OpenAI, which uses more expensive Nvidia chips to train its systems on vast swathes of data. The announcement has raised significant doubts over the future of US firms’ dominance in AI, prompting the sharp falls for Nvidia, as well as tech giants including Microsoft, Meta and Google parent Alphabet, which are all pouring billions into the technology.

The S&P 500 dropped 1.5 per cent, dragged down in large part by the fall from Nvidia. DeepSeek’s claims also affected tech stocks elsewhere, with Dutch chip making company ASML falling 7 per cent and Japan’s Softbank dropping 8.3 per cent. The FTSE 100 appeared resilient on Tuesday morning, rising 0.21% in early trading. Analysts said the announcement from DeepSeek is especially significant because it indicates that Chinese firms have innovated faster despite the US putting controls on exports of Nvidia’s most powerful chips to the country. The news marks a sharp change in fortunes for established AI companies, whose stocks have soared in value in recent years amid hopes they would reshape the world economy and deliver huge profits. Trump said he was not concerned about the breakthrough, adding that the emergence of DeepSeek could be “a positive” and a “wake-up call” for the US.

“If you could do it cheaper, if you could do it (for) less (and) get to the same end result, I think that’s a good thing for us,” he told reporters on board Air Force One. Trump also said he wanted to bring in trade tariffs that are “much bigger” than the 2.5% that some reports had suggested were favoured by incoming Treasury Secretary Scott Bessent. “I have it in my mind what it’s going to be but I won’t be setting it yet, but it’ll be enough to protect our country,” Mr Trump told reporters on Monday night. What is DeepSeek?Chinese startup DeepSeek is shaking up the global AI landscape with its latest models, claiming performance comparable to or exceeding industry-leading US models at a fraction of the cost. DeepSeek’s recent paper revealed that training its DeepSeek-V3 model required less than $6 million in computing power using Nvidia H800 chips. This figure stands in stark contrast to the billions being poured into AI development by some US companies, prompting market speculation and impacting share prices of major players like Nvidia.

Further fueling the disruption, DeepSeek’s AI Assistant, powered by DeepSeek-V3, has climbed to the top spot among free applications on Apple’s US App Store, surpassing even the popular ChatGPT. This achievement underscores the model’s capabilities and user appeal, adding weight to DeepSeek’s claims of superior performance and cost-effectiveness. The company’s rapid ascent and disruptive potential are sending shockwaves through the AI industry, challenging the established order and forcing a reassessment of investment strategies. Why is DeepSeek causing worldwide issues?The release of OpenAI’s ChatGPT in late 2022 caused a scramble among Chinese tech firms, who rushed to create their own chatbots powered by artificial intelligence.

But after the release of the first Chinese ChatGPT equivalent, made by search engine giant Baidu , there was widespread disappointment in China at the gap in AI capabilities between US and Chinese firms. The quality and cost efficiency of DeepSeek’s models have flipped this narrative on its head. The two models that have been showered with praise by Silicon Valley executives and US tech company engineers alike, DeepSeek-V3 and DeepSeek-R1, are on par with OpenAI and Meta’s most advanced models, the Chinese startup has said. They are also cheaper to use. The DeepSeek-R1, released last week, is 20 to 50 times cheaper to use than OpenAI o1 model, depending on the task, according to a post on DeepSeek’s official WeChat account. But some have publicly expressed scepticism about DeepSeek’s success story.

Scale AI CEO Alexandr Wang said during an interview with CNBC on Thursday, without providing evidence, that DeepSeek has 50,000 Nvidia H100 chips, which he claimed would not be disclosed because that would violate Washington’s export controls that ban such advanced AI chips from being sold to Chinese companies. DeepSeek did not immediately respond to a request for comment on the allegation. Bernstein analysts on Monday highlighted in a research note that DeepSeek’s total training costs for its V3 model were unknown but were much higher than the $5.58 million the startup said was used for computing power. The analysts also said the training costs of the equally-acclaimed R1 model were not disclosed.

Who is behind DeepSeek?DeepSeek is a Hangzhou-based startup whose controlling shareholder is Liang Wenfeng, co-founder of quantitative hedge fund High-Flyer, based on Chinese corporate records. Liang’s fund announced in March 2023 on its official WeChat account that it was “starting again”, going beyond trading to concentrate resources on creating a “new and independent research group, to explore the essence of AGI” (Artificial General Intelligence). DeepSeek was created later that year. ChatGPT makers OpenAI define AGI as autonomous systems that surpass humans in most economically valuable tasks.

It is unclear how much High-Flyer has invested in DeepSeek. High-Flyer has an office located in the same building as DeepSeek, and it also owns patents related to chip clusters used to train AI models, according to Chinese corporate records.