The benefits of AI manufacturing are often seen in sectors that reduce entry-level jobs.
The US is no longer leading the AI model race, says the latest Stanford AI Index report, which finds that the performance gap between the two countries’ models has “effectively closed”.
The AI Index is a program of the Stanford Institute for Human-Centered Artificial Intelligence. In its eighth edition last year, it found that while China’s AI models were faster, the US was still positioned as the clear leader in the race. However, that has changed.
Since the beginning of 2025, several Chinese models have outperformed their American counterparts, with China’s DeepSeek-R1 making its debut in February. Models from Chinese companies such as Alibaba, Zipu and MiniMax have since topped the leaderboards.
The US, however, continues to be a major supporter of AI, still producing “high-end” AI models and high-impact patents, while China leads the game in terms of volume, industrial robotics installations, citations and patent output.
Private AI investment in the US has reached nearly $285bn by 2025, with nearly 2,000 newly funded AI companies forming last year. The country also has the highest number of AI data centers.
Loss of talents
AI has undoubtedly strengthened its presence in society; Stanford reports that AI has reached mass adoption faster than the personal computer or the Internet.
Generative AI has already been adopted by more than 50pc of the population, with the numbers sitting at 61pc in Singapore, 54pc in the United Arab Emirates and around 28pc in the US.
Technology is rapidly growing in capabilities, reaching more people than ever before.
Many notable AI models released in the past year can meet or exceed human standards for PhD-level science questions, multidisciplinary thinking and competitive statistics, the report finds, creating challenging situations for job seekers. AI models built for science can outperform human scientists in many cases, it adds.
On the other hand, the report finds a connection between the declining participation rate and productivity gains.
The software development sector, which is showing the clearest signs of productivity gains, has fallen by 20pc for US workers aged 22 to 25 years. Senior positions with senior engineers are increasing in number, for now.
Despite massive investment, the US is struggling to attract global talent, with an 80pc drop in AI researchers and developers choosing to move to the country.
Responsibility takes a back seat
The report notes that responsible AI does not match the power of AI, pointing to lax security benchmarks and “spotty” reporting on benchmarks.
Documented AI events increased to 362, up from 233 in 2024. Meanwhile, recent research has found that improving AI safety can affect model accuracy, adding to the challenge of improving model safety.
The report also touches on the sovereignty of AI, calling it “a defining feature of national policies”.
The EU, for the first time, launched a Continental Action Plan for AI last April, promising to develop AI infrastructure and reduce dependence on its technical needs.
However, new open source developments – in particular, OpenClaw – are helping to redistribute the stake in the AI race.
Technology firms that use open source models are widely accessible by creating their own versions of OpenClaw with enhanced security.
Don’t miss out on the information you need to succeed. Sign up for Daily BriefSilicon Republic’s digest of must-know sci-tech news.
