When geopolitics meets innovation – China’s DeepSeek raises stakes in AI arms race

"DeepSeek’s arrival on the AI scene in late January stunned investors and prompted a dramatic reassessment of the perceived dominance of US technology."

The development of a comparable model that is apparently more efficient and available at a fraction of the cost of its competitors has disrupted the prevailing narrative about the future of AI. By challenging the view that bigger is better when it comes to AI, this episode has fuelled doubts about the valuations of US tech giants, their growth prospects and planned investments in AI.

According to Jeffrey Lin, M&G Investments’ Head of Thematic Technology Equities, the implication of DeepSeek, and the reason Nvidia bore the brunt of the sell-off, is that AI models in future might require less computational power.  

“Investors anticipated the efficiency gains would mean companies won’t need to commit as much investment as previously announced to integrate AI into their products or processes,” he says.

“The figure for DeepSeek’s spending may turn out to be too low, however, as they may have used some prior work before building their model. While there is a lack of clarity around the exact cost, in terms of performance DeepSeek appears to have been validated.” 
 

AI arms race

DeepSeek’s cutting-edge capabilities were also surprising as it was widely assumed that US tech firms were ahead of their Chinese counterparts in the AI race. President Trump described the DeepSeek development as a “wake-up call” for US companies in an area where the US is determined to be the global leader.  

Emphasising that AI is likely to be a priority for his administration, the president announced a $500 billion investment in AI infrastructure, the Stargate Project, which he said was “the largest AI infrastructure project by far in history”.³

Being at the forefront of this nascent technology is now a worldwide issue. At the AI Action Summit in Paris in February, the European Union (EU) set out its ambition to compete in the AI race, pledging €200 billion to accelerate innovation in the space⁴. French President Emmanuel Macron also unveiled €109 billion of investments, directed towards developing AI technology in France, including the construction of data centres⁵. France-based Mistral AI has developed high-performing open source AI models and is seen as one of Europe’s best chances of keeping up with the US and China.

Dominance of the AI landscape

Number of notable machine learning models by geographic area, 2023

Source: The AI Index 2024 Annual Report, Stanford University.

AI is a core element of the so-called “fourth industrial revolution” due to the transformative impact it could have on the way we live and work. It is increasingly regarded as a critical factor for national success across many global economies in the coming years. AI leadership could provide substantial economic benefits through productivity gains as well as creating new markets. It could also result in improvements in healthcare and education and other aspects of society. 

Another powerful reason why countries are so keen to be an AI leader is the potential advantages for national security through intelligence and weapons. In an increasingly unstable world, and with more technological warfare, cutting-edge military capabilities are becoming more critical.  

Given the wide-ranging and significant advantages that AI could bring, it is not surprising that there is a global competition to be at the front of the pack in this field. 

Despite the rivalry, there is a recognition that this powerful emerging technology could have negative consequences as well as being a force for good. The loss of jobs through automation, the possibility of bias and discrimination, and security concerns are often cited as risks that could flow from the development of AI. 

As a result, there have been efforts, such as Europe’s AI Action Summit, as well as the AI Safety Summit in the UK in 2023, to encourage collaboration and establish international norms to harness the benefits and mitigate the risks of this technology.  

While this international approach is encouraging, it appears that the Trump administration is taking a different stance. At the Paris summit Vice President JD Vance explained that the US favours less regulation of AI and will “make every effort to encourage pro-growth AI policies.” One of Trump’s first acts in office was to revoke President Biden’s order that required firms to focus on AI safety.

This desire to push ahead arguably illustrates that the world’s two largest economies are in a race to be the AI leader, which can be regarded as an extension of their broader struggle for economic and strategic dominance on the world stage. 
 

Innovation in China

It is hard to say at this stage who will eventually become the dominant player in AI but the US has tried to slow China’s progress. It has prevented Nvidia exporting its most up-to-date chips to China, and restrictions have been placed on chip-making equipment reaching China. 

The fact that a Chinese company was able to produce a cutting-edge model, despite being denied access to the latest US hardware, was one of the biggest surprises about the DeepSeek episode. However, this shouldn’t come as such as shock.

“China has many talented computer scientists and the fact they have been able to develop this product with much less computational power demonstrates their capabilities,” explains Lin. “The attention has been on US firms, but DeepSeek arguably validates China as a major force in AI globally.”

DeepSeek’s initial success highlights how China’s economy has changed in recent years, with a new emphasis on innovation-led development. The 'Made In China 2025' policy was launched in 2015 with the aim of transforming the Asian superpower’s manufacturing sector and shifting towards high-tech industries. 

This was followed in 2023 when President Xi Jinping introduced the concept of “new quality productive forces”. The Chinese government’s new focus on driving economic growth through innovation and technological advancement, rather than the traditional model of property and infrastructure, has become increasingly evident over the years.

As a result of these policies, China today is a highly innovative economy that is at the forefront of many industries. The country may have a reputation as a manufacturer of low-quality goods but 'Made In China' is also now synonymous with technologically advanced products. 

Last year, China filed the largest number of patents in the world: 1.7 million compared to 600,000 in the US. There are also more than 6,000 PhDs in STEM (science, technology, engineering and maths) coming out of China’s universities each month, double the figure for the US.⁶  

This tech-focused environment is helping China become a market leader in areas ranging from electric vehicles (EVs) to solar panels and battery technology. At the end of last year, Chinese car maker BYD overtook Tesla to become the world’s largest producer of electric vehicles.⁷
 

Is AI fuelling optimism?

As DeepSeek demonstrates, US efforts to slow China’s progress appear to have inadvertently driven technological developments and innovation. Chinese firms have thrived without advanced US chips and software. Phone maker Huawei has produced a competitive smartphone with an advanced domestically-produced processor, for example. It has also developed its own operating system, as it was cut off from Google’s Android. 

China is home to a myriad of technologically advanced companies including several tech giants such as Alibaba, Baidu and Tencent that are comparable to the Magnificent 7 in the US. Many of these have well-established and competitive AI models. 

“China is one of the most digitally advanced economies globally, whether measured by high internet and smartphone penetration rates or world-leading digitally integrated supply chains,” remarks David Perrett, Co-Head of Asia Pacific Equities. “This technological advancement offers a real competitive edge to companies, many of which, in our view, trade at attractive levels of valuation.” 

One consequence of DeepSeek’s emergence is that investors have started to reassess Chinese AI-related investments. Technology stocks including Alibaba and Baidu rallied after the breakthrough amid growing optimism about the development of AI in the country.

Alibaba’s share price gained when Apple selected the firm as its partner for its AI services in China and BYD’s share price received a boost when it announced it was teaming up with DeepSeek to incorporate AI into its vehicles.  

For the past couple of years, concerns about China’s economic outlook, including its troubled property sector and domestic consumption, have led investors to steer clear of China’s stock market. But growing appreciation of the country’s AI capabilities and the prospect of widespread AI adoption appears to have helped restore some confidence in the country’s equity market. Given how well US tech stocks performed during the AI buzz of the past two years, could Chinese AI and tech stocks experience a similar rally?

“Despite the malaise in the Chinese stock market in recent years, we have remained excited about the technological advances we have seen coming from the country,” says Perrett. 

One clear lesson from DeepSeek is that China’s investment in the future of technology is bearing fruit. Could it also help revive the country’s stock market and economy? Perrett is optimistic: “We believe with our disciplined approach to bottom-up research, we can find differentiated ideas from across various sectors, which can help drive returns for our clients.”
 

Artificial intelligence is a multi-decade opportunity

While the rivalry between the US and China will likely have far-reaching consequences for the global economy, trade and geopolitics, in the near-term DeepSeek could have some significant implications for the evolution of AI. 

Investors’ initial panic suggests that it undermined the optimistic case for big tech’s growth and their investments. An alternative view is that a cheaper, more efficient AI technology could actually represent a catalyst for wider adoption and deployment of AI. Companies that might have been cautious about the costs associated with implementing AI may be relieved that it is available more cheaply.

Lin is optimistic that DeepSeek could have a positive effect. “In the history of computing, whenever computers become more powerful and new applications for computing emerge, the addressable market increases,” he observes.

“For companies that want to incorporate AI into their products, particularly enterprise software companies, it's becoming less capital intensive and it's moving faster. They can add these enhancements to software more quickly so the end market actually grows as well,” he says.

Despite the concerns that DeepSeek’s efficiency gains might mean reduced demand for Nvidia’s chips, Lin believes that the increasing pace of innovation and growing addressable market will ensure that demand remains robust. 

There may be a shift in focus from training foundational models towards applications and services, but overall demand for AI technology is likely to continue growing. Big US tech firms certainly remain committed to spending on data centres and chips, with Alphabet (parent company of Google), Amazon, Meta and Microsoft forecast to invest $320 billion on AI this year⁸.

Lin sees AI as a multi-decade opportunity that is still in its infancy. “It began with image recognition and predictive analytics and now generative AI is the current big use case,” he explains. “Beyond this there are other AI opportunities such as agentic AI, where computers start to be able to reason like humans do. There’s also robotics – industrial robots and maybe even personal robots in the future – as well as self-driving cars.” 

In his view, the evolution of AI is creating different investment opportunities. “To begin with you have the enablers, the companies that provide foundational technology for AI, such as semiconductor firms like Nvidia. Then there are the providers: companies, particularly in the enterprise software area, which can use the technology to offer AI-enhanced products and services, and importantly charge extra for them. Finally, there are what we call the beneficiaries, which can use AI internally to grow their business or just become much more operationally efficient.” 

In the two and a half years since ChatGPT created the buzz around generative AI, most of the focus has been on the enablers, notably Nvidia. The fact that AI has yet to become a feature of everyday life for many people prompted concerns that it might have been overhyped. However, there are signs that the technology is now moving towards the providers. If DeepSeek provides a further boost to adoption, this progress could gather momentum in the near future. 
 

The age of artificial intelligence

The age of AI is arguably just beginning and the landscape is evolving fast. DeepSeek disrupted the prevailing narrative about AI and, in Lin’s view, may become one of the important milestones in the evolution of the technology. “Ultimately, we think it's a sign of acceleration of the development for AI. We can potentially get to new AI applications sooner rather than later,” he suggests. 

The US and China appear to be leading the AI arms race but other countries are determined to compete to ensure that they can benefit from the significant advantages associated with the technology. The race may prove to be more of a marathon than a sprint but for Lin there will be plenty of long-term investment opportunities as the pace of innovation and deployment accelerates in the years ahead.