In January 2025, the emergence of Chinese AI DeepSeek shook the global tech landscape and caused many US tech stocks to plummet, with US President Donald Trump dubbing it a “wakeup call” for US tech companies
Founded in 2023 by Liang Wenfeng and headquartered in Hangzhou, DeepSeek specialises in developing open-source large language models (LLMs) – advanced AI models trained on vast amounts of data to understand and generate everything from poetry to Java code. Its flagship model, DeepSeek-R1, has garnered significant attention for its performance and cost-efficiency, challenging established Western AIs like OpenAI’s ChatGPT and Google Gemini.
A new AI contender
DeepSeek’s chatbot offers capabilities comparable to leading platforms like ChatGPT but distinguishes itself through its development efficiency. The company claims the model was trained at a cost of approximately $6 million, a stark contrast to the estimated $100 million expenditure for OpenAI’s GPT-4 in 2023. Moreover, DeepSeek-R1 requires only a tenth of the computing power of similar models, highlighting its resource efficiency.
Early users have found that the model performs as well as ChatGPT and Gemini, although many have raised questions about censorship (which will be explored further below). Performance and news headlines have brought a lot of attention as a result, and DeepSeek’s first free chatbot app for iOS and Android platforms surpassed ChatGPT as the most-downloaded free application on the US iOS App Store on 27 January.
Market disruption and economic implications
The swift rise of DeepSeek has had profound effects on global markets. DeepSeek’s open source model and lower development and computing costs undercut a common belief in Silicon Valley that AI can only advance with the input of huge budgets and top-tier chips.
As a result, shares of major technology companies, particularly those heavily invested in AI infrastructure, experienced sharp declines when markets opened on Monday, 27 January. For example, chip manufacturer Nvidia saw its stock price drop by 18% over concerns that DeepSeek’s efficient models could reduce the demand for chips, thereby impacting Nvidia’s future revenue streams.
Strategic implications for countries looking for AI supremacy
DeepSeek’s emergence underscores China’s rapid progress in AI. This has raised concerns about the effectiveness of bans on advanced chip and technology exports to China and prompted discussions about the need for strategic investments to maintain a competitive edge.
The company’s success also challenges long-held stereotypes about Chinese innovation, demonstrating that China is capable of being a leader rather than a follower and producing high-performance, cost-effective AI solutions. Chinese media have widely praised DeepSeek for its small yet formidable team, primarily comprised of young graduates from China’s top universities who have been deeply immersed in the tech field from a young age.
And it is not the only Chinese company purporting to be breaking new ground in the AI field. On 29 January, Alibaba (owner of Taobao, Tmall and Alipay, among others) announced a new version of its Qwen 2.5 AI model that it claims surpasses DeepSeek, OpenAI and Meta’s latest models.
Ethical and regulatory considerations
Despite its achievements, DeepSeek has faced scrutiny over data privacy and censorship concerns. The company’s models reportedly adhere to Chinese censorship laws, avoiding politically sensitive topics, which has raised questions about the ethical implications of such restrictions.
Moreover, there have been allegations that DeepSeek illicitly used OpenAI’s models to train its own through a technique called “distillation”, potentially infringing on intellectual property rights. This has caused some to question DeepSeek’s claims about how it produced its model so cheaply, although it should be noted that models like ChatGPT have also been criticised for infringing on intellectual property rights.
Finally, in the wake of the US Supreme Court upholding a law that could ban TikTok in the US over national security concerns (since being pushed back by an executive order from Trump), some have raised similar questions about DeepSeek’s collection, use and storage of data. As The Guardian reports, DeepSeek’s privacy policy states that the personal information it collects is held on secure servers in China.