As large language models advance at a rapid pace, artificial intelligence is increasingly being integrated into a wide range of workplaces, including offices, software development, and design studios. One of the most pressing questions in this wave of technological change is how many jobs AI will replace—and how many new employment opportunities it will create.

According to a study released in March 2026 by the International Labour Organization and the World Bank, generative AI is reshaping the global labor market, but the opportunities and disruptions it brings are not evenly distributed across countries. Some developing economies may experience employment turbulence sooner than others. Meanwhile, new occupational demands are also emerging. Data from the recruitment platform Maimai Gaopin Talent Intelligence shows that from January to May 2026, the number of newly posted campus recruitment positions in AI grew by 47.3% year-on-year.

In a recent discussion on the topic, Yao Yang, Dean of the Lingang Advanced Finance Institute at Shanghai University of Finance and Economics and a visiting professor at Peking University, offered his assessment. He emphasized the importance of distinguishing between the short-term and long-term effects of AI on employment. "In the short term, I believe more jobs will be eliminated; in the long run, more new positions will certainly be created," Yao said, adding that, more precisely, AI will generate entirely different categories of work.

To illustrate the forms these new roles might take, Yao drew an analogy with the development of the internet industry. "Over time, you will find that many new industries will emerge around AI, just like the internet," he said. Around the year 2000, the internet sector entered a phase of rapid growth. Over the following two decades, a new business ecosystem gradually formed around it—including e-commerce, mobile payments, online marketing, on-demand delivery, and platform services—giving rise to countless professions and jobs that did not exist before.

Yao pointed out that while the internet disrupted traditional brick-and-mortar retail, it simultaneously gave birth to new business models and employment ecosystems. "The internet created an ecosystem. Over the past 20 years, that ecosystem has gradually been built up. Of course, it destroyed a number of physical stores, but new formats have also emerged," he said.

In Yao's view, AI is still in its early stages of development, and its impact may go far beyond simply replacing existing jobs. As the technology enters more industries, a new industrial ecosystem will form around models, computing power, and applications. In the future, we may see types of work and economic structures that are difficult to foresee today. "What kind of economic form AI will ultimately create, we still do not know at this point," Yao concluded.