In February 2026, there was a significant turbulence in the global technology stock market. The software sector faced massive sell-offs, with market capitalization evaporating over trillions of dollars, leading to a 30% to 50% drop in stock prices of some leading companies. At the same time, semiconductor and hardware companies witnessed strong stock price increases, resulting in a significant rise in market capitalization. Experts pointed out that this “software-hardware differentiation” stemmed from the disruptive impact of AI technology on traditional software business models, leading to a significant reshaping of the industry. They also urged government intervention to regulate the AI industry, warning of alarming consequences if left unchecked.
On Friday, February 6, the Dow Jones Industrial Average broke the 50,000-point mark for the first time. After several days of sharp declines in tech stocks, there was a noticeable rebound that day, driving a strong overall market surge. The Dow soared over 1,100 points, a gain of about 2.3%, reaching a peak of 50,015.07 points during the trading session, setting a new all-time high and marking the first time the index crossed the 50,000-point threshold. This surge also helped the S&P 500 index turn positive for the first time in 2026.
According to a report by Hong Kong’s “Sing Tao Daily” on the 5th, both the Hong Kong and US stock markets recently experienced mini stock crashes, primarily dragged down by technology software stocks. For over two decades, software stocks were considered a high-growth industry, while hardware manufacturers were often deemed to have limited profits. However, a recent market reversal saw an uptick in the hardware sector while software stocks continued to decline.
In the US market, tech giants like Microsoft, Oracle, Adobe, and Salesforce led the decline, with each seeing a drop of 30% to 50% from their peaks. Bloomberg reported that within just two days, the stocks, bonds, and loans of Silicon Valley companies evaporated by billions of dollars. The software sector took the hardest hit, with software stocks tracked by iShares ETF shedding over a trillion dollars in market value over the past week.
The US stock market has recently shown a divergent trend. While most S&P 500 components rose, the downturn in large tech stocks dragged down the three major indices. Goldman Sachs’ basket of US software stocks also experienced a significant decline.
In the Hong Kong market, Tencent and Alibaba have seen recent declines, with the former dropping by 10% since last Friday, and the latter also falling by over 8% during the same period. Some mid-sized software stocks saw even sharper declines, with Kingdee, Meitu, and JS Tank plummeting by over 10% in a single day.
Market analysis widely believes that the launch of new AI products has intensified investors’ concerns about the prospects of the software industry.
An ex-Shanghai stock analyst, Bu Qingsong, pointed out two main reasons for the decline in software stocks during an interview with Dajiyuan. On the one hand, large tech companies are significantly increasing their AI-related capital expenditures, such as purchasing chips and building data centers, thus squeezing profit margins in the short term. On the other hand, the technological impact brought about by the AI wave is changing the competitive landscape of the software industry.
Bu Qingsong stated that at a deeper level, traditional software companies are facing severe strategic dilemmas. Investors are increasingly worried about the “AI bubble,” particularly with mega-companies like Microsoft and Oracle needing to invest hundreds of billions of dollars in acquiring chips and constructing servers to build AI capabilities. The returns on these investments remain uncertain, leading to a continuous “burn” of funds, which could potentially replicate the bursting of the Internet bubble in the early 2000s.
Moreover, some investors believe that as AI programs become more powerful, anyone can design sophisticated software without needing to understand programming, merely through natural language interaction. Bu Qingsong pointed out that core functions of some traditional software products may be directly completed by AI models in the future, weakening the existing business moats.
Bu Qingsong expressed that in the current narrative, traditional software companies are in a dilemma. “If AI is indeed a ‘bubble that will burst sooner or later,’ the billions of dollars invested by software giants will go down the drain. If AI is not a bubble, and it becomes omnipotent in the future, software companies may lose a significant amount of business when that time comes. The industry has entered a stage of either transformation or elimination.”
In contrast to the software stocks, semiconductor and hardware company stocks have generally risen.
From Nvidia and TSMC to ASML, several chip and equipment manufacturers set new stock price highs. Over the past year, the stock prices of Samsung Electronics and SK Hynix in South Korea surged by approximately 220% and 370%, respectively, with the total market capitalization of the two companies exceeding $1.2 trillion, surpassing the combined market value of Tencent and Alibaba at around $1.05 trillion.
Analysts liken this phenomenon to “selling shovels during a gold rush.” Even if there’s a risk of an AI industry bubble in the future, related companies still need to purchase a large number of chips and server equipment, ensuring more certainty in short-term hardware demand.
According to a report by Bloomberg, the software stock crash is not limited to US-listed companies but has expanded to Wall Street supporters, including lending institutions and private equity firms.
Bu Qingsong believes that the current turmoil appears to be more of an adjustment during the rise of the AI industry, rather than a bubble bursting. He stated that the AI era has just begun, but the differentiation between companies will become more apparent.
He said, “The current divergence of tech stocks reflects AI reconceptualizing the entire industry chain landscape. The hardware sector benefits from genuine demand, while the software industry faces pressure to reshape its business model. With the continuous advancement of AI technology, the tech industry will enter a new round of reshuffling. The success or failure of transitioning will be crucial to the survival of software companies in the future.”
Some tech experts attribute the current market volatility to the entry of AI agents into the application layer.
A Chinese hardware engineer in California, Yu Zhiming (pseudonym), told Dajiyuan that this downturn differs from previous AI bubble panics and stems from the substantial disruption that AI is bringing to various industries’ business models.
Specifically, last Friday, Anthropic released the Claude Cowork plugin aiming to automate workflows in fields like law, finance, marketing, and data analysis, which is seen as a significant signal of AI agents beginning to invade the “application layer.” He believes this implies potential disruptions to vertical software models.
Yu Zhiming pointed out, “Traditional software companies need to adapt quickly, build an AI ecosystem to maintain competitiveness, or else their market value will continue to evaporate, even facing bankruptcy—similar to how digital cameras replaced film cameras, traditional manufacturers might face elimination if they don’t innovate.”
Regarding hardware companies, Yu Zhiming acknowledged that they are indeed beneficiaries of the AI prosperity in the short term but face challenges in the long run. As AI models optimize, such as with smaller models or edge computing advancements, the growth rate in hardware demand might slow down. Increased competition could also compress profit margins, and if the AI investment bubble bursts, hardware manufacturers would similarly be impacted.
However, he emphasized that there is still a high dependence between software and hardware. The main clients of hardware companies are still software and cloud service providers, so a decrease in demand from the latter could also affect the hardware industry.
Additionally, Yu Zhiming expressed concerns about the societal impacts of AI.
A few years ago, Tesla CEO Elon Musk and other tech leaders and researchers urged all AI labs globally to halt the development of the most advanced systems. The letter listed a series of risks that AI could pose to humanity. The letter questioned whether we should develop non-human thinking that could eventually surpass, outsmart, eliminate, and replace us. It raised concerns about the potential loss of control over civilization.
Yu Zhiming called for government intervention to regulate the AI industry, warning of alarming consequences if its development continues unchecked. Technological advancements have historically liberated people from tasks, but full AI replacement could lead humanity to live a “scientific pig-rearing” life: exchanging free feed for tasks like playing games to provide computing power.