How China responds to AI’s impact on employment
A technician at a high-tech manufacturing company inspects in-house developed electronic component protection products. Photo: IC PHOTO
Vowing to advance the AI Plus initiative, the 2025 government work report proposes to support the extensive application of large-scale AI models and to vigorously develop new-generation intelligent terminals and smart manufacturing equipment. Like previous technological revolutions, AI as a general-purpose technology will see extensive application across sectors, reshaping economic and industrial structures and prompting structural adjustments in the labor market. Against this backdrop, what kind of shocks and structural impacts will AI technology bring to China’s labor market, and how should we respond?
Dual effects of job substitution and creation
AI refers to technologies capable of performing tasks that display human-like intelligence. With the application of machine learning and neural networks, AI has achieved breakthrough progress in recent years. These technologies excel in areas such as image recognition, speech translation, autonomous driving, intelligent customer service, and generative AI based on large models, enabling the automation of tasks once considered exclusive to human labor.
In this new technological paradigm, which jobs are likely to be replaced first? Research shows that repetitive tasks and those with well-defined procedures are most readily encoded and thus automated. As AI technologies advance and improve, this substitution has gradually expanded—from manual to cognitive labor, and from low-skilled to higher-skilled jobs. Earlier, industrial robots and robotic arms replaced assembly-line workers in manufacturing, while vending machines and ATMs displaced sales clerks and bank tellers. More recently, substitution effects have also appeared in knowledge- and technology-intensive industries: AI has proven effective in medical imaging diagnostics and searching and generating legal documents, and has begun to displace jobs such as radiologists and paralegals.
In practice, whether AI substitution occurs depends not only on the technical feasibility of encoding tasks, but also on cost considerations. As the price of digital equipment falls over time, businesses will increasingly adopt AI whenever it offers cost savings.
At present, specialized AI is advancing rapidly and, for certain specific tasks, can even outperform humans. Growing exponentially, it is constantly pushing the boundaries of how broadly and deeply it can substitute for human work. However, AI still struggles to accomplish the full range of intellectual tasks humans can perform. The future of artificial general intelligence remains uncertain: its development is prohibitively costly, and whether it can ever truly be realized is still debated. In particular, where production tasks are complex and require interaction, decision-making, or reasoning, human labor retains a clear advantage.
While AI displaces workers in certain roles, it simultaneously gives rise to many new types of jobs. Occupations are a typical form of social division of labor, and different professions are often complementary. Once AI replaces simple and routine tasks, workers can concentrate on areas and roles that AI cannot replace, applying uniquely human abstract thinking, creativity, and judgment.
Studies find that AI complements certain non-routine tasks. Non-routine cognitive tasks require cognitive and analytical abilities, as seen in the work of programmers, analysts, and engineers. More broadly, non-routine interactive tasks rely on professional experience and interpersonal communication skills, such as those performed by doctors, teachers, and lawyers. These non-routine tasks are difficult to codify and automate, and in such areas AI serves as a complementary aid, enhancing workers’ productivity.
In addition, the spread of AI creates demand for people to train, operate, and maintain these systems, giving rise to new professions such as AI trainers, algorithm engineers, and multimedia operations specialists.
Structural impact of AI on labor market
AI reshapes the occupational structure by transforming the division of labor. First, AI’s substitution of human labor eliminates certain traditional occupations. Once the primary tasks of a job can be performed by AI, intelligent machines—embodying capital—take over that segment of the division of labor, and the human factor exits. Jobs characterized by simple tasks and low skill requirements are the first to be replaced and to disappear from the social stage, such as assembly-line workers, ticket clerks, cashiers, and office clerks.
Second, the application of AI continuously generates new positions, professions, and industries. To begin with, AI-driven applications foster new productive forces, raising social productivity and boosting household incomes, which in turn expands high-income-elastic service industries and creates employment in fields such as healthcare, education, and leisure. Furthermore, AI research and application rely on the support of technology-intensive industries such as information technology, data science, and semiconductor manufacturing, thereby driving demand for highly skilled professionals in these fields.
Finally, AI’s role in upgrading and transforming traditional industries has created a large number of new occupations and positions. According to the World Economic Forum’s Future of Jobs Report 2025, demand for roles such as big data specialists, fintech engineers, and AI experts will grow by more than 80% in the next five years. Between 2019 and 2021, China’s Ministry of Human Resources and Social Security announced more than 50 new professions. These included AI-related roles such as cloud computing engineers and big data specialists, industrial robot operators and maintenance technicians created through industrial upgrading, and emerging positions in modern service industries such as internet marketing specialists and online learning facilitators.
Skills refer to the competencies workers must possess to perform occupational tasks, and AI is reshaping the core skill requirements of the labor market. First, the skill requirements and task structures of existing occupations are changing, with an increasing emphasis on non-routine skills. For example, with the widespread adoption of personal computers and office software, secretaries no longer need to master mechanical typing but must be proficient in digital literacy, software applications, teamwork, and organizational communication.
Second, as AI replaces routine jobs and generates new ones, the division of labor grows more specialized. Workers now require not only professional skills specific to their field but also the ability to collaborate across disciplines. This increases demand for cognitive and social skills—the two “soft skills” that AI cannot easily replace. In recent years, demand and wage premiums for these skills have both been rising. Looking ahead, cultivating AI literacy and the ability to apply AI will be equally critical. Workers will need to leverage human–machine collaboration, focus on tasks beyond AI’s reach, and continuously learn and expand their skill sets to reduce the risk of displacement.
Taking active measures to address AI’s impact
General Secretary of the CPC Central Committee Xi Jinping said at the 10th meeting of the Central Committee for Financial and Economic Affairs in August 2021 that “A new round of technological revolution and industrial transformation is a powerful driving force for economic development and imposes a significant impact on employment and income distribution, including some negative effects that must be effectively solved.”
AI exerts diverse structural effects on the labor market. In the long run, its role in upgrading traditional industries and creating new occupations will support higher-quality employment. In the short term, however, it also creates frictions within the labor market, leading to unemployment and efficiency losses. Government departments should adopt a well-coordinated mix of diversified and structural policies to effectively mitigate the short-term shocks of AI on the labor market.
From the perspective of occupational structure, workers engaged in routine jobs are most vulnerable to AI’s negative impacts. As demand for such workers declines, they may be forced to leave their jobs, resulting in income loss and reduced welfare. Low-skilled workers and middle-aged or older workers form the bulk of this group, and for them, acquiring new skills and finding new employment often requires more time. During this transitional period, relevant authorities should provide regionally tailored social protection measures and deliver targeted skills training for disadvantaged groups in the labor market.
From the perspective of skills structure, skill development is a “slow variable.” The education system should play a proactive role in shaping students’ competencies and skills. Higher education and vocational training should be dynamically optimized in both curricula and specializations to meet the needs of new professions and skills in the AI era. Greater emphasis should be placed on cultivating students’ critical and innovative thinking, as well as “soft skills” such as communication and teamwork. At the same time, workers themselves should take the initiative to learn and use AI tools, developing the capacity to collaborate with machines in solving complex problems. Society as a whole must foster a culture of agile and lifelong learning, helping workers adapt to the exponential development and rapid iteration of AI technologies.
While focusing on AI’s disruptive effects on employment, equal attention should be given to leveraging AI to improve job quality. In today’s labor market, practical skills may matter more than formal academic qualifications, and the rise of digital platforms has enhanced the efficiency of skill-to-job matching. By improving worker–job matching, digital platforms can not only raise workers’ incomes but also boost enterprise productivity.
Chen Lin is a professor from the School of Economics and Management at East China Normal University; Gao Yuepeng is from the School of Economics at Shanghai University of Finance and Economics.
Editor:Yu Hui
Copyright©2023 CSSN All Rights Reserved