Summary of Key Points
In the past few years, office buildings in China's core cities have faced pressure due to an oversupply and the reduction in office space by traditional enterprises (such as financial and internet giants). However, the explosive growth of the AI and semiconductor industries in the past year has become a "key factor" that is changing this situation. AI companies are not only renting large amounts of premium office space but also purchasing entire buildings, leading to a differentiation in the office market—specific areas have become "overheated" due to AI demand. The impact of AI is even extending to the residential market. Nevertheless, there is still caution within the industry regarding whether AI will be able to sustain this positive trend for the long term.
1. AI Companies Becoming the New Drivers of the Office Market: From Renting to Buying
In the past, the recruitment of office tenants mainly relied on financial institutions and internet giants; now, promotional materials highlight "AI industry ecosystems" and "large model clusters." The demand from AI companies is distinct: they require large spaces, quick decision-making processes, and high standards for buildings. For example, the "Yunjishangpu" project in Yangpu District, Shanghai, was immediately rented out by ByteDance's "DouBao" subsidiary for 40,000 square meters (equivalent to half a building), nearly filling the rental capacity of this newly delivered property. Another example is Zhipu AI, which purchased the Diamond Building in Zhongguancun, Beijing, for 361 million yuan to use as its headquarters, becoming neighbors of Baidu and Tencent. This move serves both office needs and asset optimization through property value appreciation.
This trend is not limited to Shanghai; AI companies in Shenzhen's Nanshan Science Park and Hangzhou's YunGu have also moved from incubators into premium office buildings, with a noticeable increase in demand.
2. Market Differentiation: Hot Areas for AI, Stable Areas for Traditional Industries
Despite an overall oversupply of office space in core cities, different areas are experiencing vastly contrasting fortunes:
- Traditional Financial and Business Zones: Areas like Lujiazui in Shanghai see stable demand with little fluctuation.
- AI/Computing Power/Chip-Focused Zones: Regions such as the Binjiang West Shore Digital Belt in Xuhui and Zhangjiang Moduli Community have seen "local overheating." These areas share common characteristics: government support through policies, industry clustering, and significant customer demand, leading to concentrated rental transactions. For instance, Zhangjiang Moduli Community is tailored for large model companies, with subsidies from the government and a concentration of chip and computing power companies, resulting in faster leasing turnover compared to other areas.
3. AI's Impact on the Residential Market
The expansion of the AI industry is not isolated; it has created a "chain reaction" that affects both commercial and residential real estate:
- Office Buildings: AI companies' increased rent payments have led to a rise in the "net absorption" of office space in first-tier cities (the difference between new rentals and vacancies). The vacancy rates in Shenzhen and Hangzhou's technology zones have dropped from 30% to 20%, indicating a decrease in available offices.
- Residential Market: The higher salaries and equity returns for AI company executives and key technical personnel have boosted residential sales in these cities. A UBS report suggests that the recent improvement in the real estate market is driven by actual demand from the AI industry, with industrial profits increasing by 15.5% in the first quarter of 2026, with AI-related sectors being a major contributor.
4. Cautionary Views: The Impact of AI Still Needs to Be Observed
Despite the positive trends, many industry insiders remain cautious:
- Small Employee Base: Compared to the previous internet boom, AI companies have higher valuations but smaller employee numbers. For example, an AI company valued at tens of billions may employ only a few hundred people, whereas previous internet companies often rented entire buildings with thousands of employees.
- Delayed Impact: Most AI companies are still in the financing and recruitment phases, so the full realization of their office demand may take time. Some advisors believe that we are only seeing the initial signs of growth and have not yet reached a "super tenant era."
Conclusion
The AI industry has indeed provided a boost to the sluggish office market, but its ability to sustain this positive trend depends on the profitability of AI companies, the pace of employee expansion, and the stability of related policies. For now, it is a variable factor rather than a definitive solution. The changes that ordinary people can observe include higher office rents in areas with many tech companies and stable housing prices in those surrounding them.