AI in the Workplace: How Employees Really Feel
As artificial intelligence (AI) continues to reshape industries, a growing body of research is turning its attention from technological capabilities to human responses. While headlines often emphasize job displacement and automation fears, a new study offers a more nuanced and surprisingly optimistic picture of how employees are actually experiencing the rise of AI in their workplaces.
Conducted by researchers from Beijing Union University’s School of Management, the study surveyed 490 employees across 98 companies actively deploying AI technologies. The findings, published in the journal Enterprise Human Resource Management, challenge widespread assumptions about AI-induced anxiety and instead reveal a workforce that is largely resilient, engaged, and even energized by the integration of intelligent systems.
The research, led by Guo Juan, Zhu Xiaomei, Li Ziying, and Du Hui, moves beyond macroeconomic forecasts to examine the psychological undercurrents of the AI revolution. Rather than treating employees as passive victims of technological change, the study investigates the emotional and cognitive responses—ranging from psychological well-being and anxiety to a sense of vitality at work—across different industries, roles, age groups, and educational backgrounds.
One of the most striking findings is the overall positive psychological state of employees in AI-intensive environments. Contrary to the narrative of widespread fear and insecurity, the data shows that employees report high levels of psychological well-being and work thriving, with relatively low levels of anxiety. On a five-point scale, psychological well-being averaged 3.74, work thriving reached 4.17, and anxiety measured just 2.07—indicating that most employees feel more positive than negative about their work lives in the age of AI.
This optimism is not uniform, however. The study reveals significant differences based on industry, role, age, and education, offering a detailed map of where psychological risks and opportunities lie.
Industry Matters: Manufacturing Feels the Pressure, Transportation Thrives
The research identifies clear patterns across sectors. Employees in the manufacturing industry reported the highest levels of anxiety (2.25), likely due to the visible presence of industrial robots replacing manual labor. Similarly, those in consumer services—where service robots are increasingly common—also showed elevated anxiety, suggesting that direct job substitution creates palpable stress.
In contrast, transportation industry employees reported the lowest anxiety (1.51) and some of the highest levels of work thriving (4.33). This may be because AI applications in transportation, such as route optimization and fleet management, are currently augmenting rather than replacing human roles. Employees in this sector may perceive AI as a tool that enhances efficiency and safety rather than a threat to their livelihoods.
Information technology (IT) employees, despite working at the forefront of AI development, showed a more complex profile. While their work thriving was high (4.13), their anxiety levels were also elevated (2.12), and their psychological well-being was slightly below average (3.68). This suggests that even those building AI systems are not immune to the pressures of rapid technological change. The constant need to learn, adapt, and stay ahead of the curve can be both energizing and exhausting.
Scientific research and cultural education sectors showed more balanced outcomes, with psychological states close to the overall average. This may reflect the slower pace of AI integration in these fields, allowing employees more time to adjust.
The Paradox of Experience: Longer AI Exposure, Lower Morale?
One of the most counterintuitive findings concerns the duration of AI adoption within companies. Intuitively, one might expect that as organizations gain experience with AI, employees would become more comfortable and confident. However, the data suggests the opposite trend.
Employees in companies that have used AI for over five years reported slightly higher anxiety (2.21) and lower work thriving (4.14) compared to those in organizations with shorter AI experience. While these differences were not statistically significant, the overall trend points to a potential “adaptation fatigue”—a phenomenon where the initial excitement of innovation gives way to the sustained pressure of continuous learning and performance expectations.
This finding underscores a critical insight: the psychological impact of AI is not a one-time adjustment but an ongoing process. As AI systems evolve, so do the skills required to work with them. Employees may face repeated cycles of retraining, role redefinition, and performance evaluation, which can erode morale over time if not managed carefully.
Age and Attitude: The Generational Divide in AI Acceptance
Age emerged as a significant factor in shaping employee responses. Younger workers, particularly those aged 26 to 30, reported the highest levels of psychological well-being (3.83) and the lowest anxiety (1.98). This cohort, largely composed of digital natives, appears to view AI as a natural extension of their work environment—an opportunity for growth rather than a threat.
In contrast, employees aged 41 and above reported the lowest well-being (3.46) and among the highest anxiety levels (2.19). This age group may face unique challenges, including concerns about skill obsolescence, reduced career mobility, and the difficulty of mastering new technologies later in their careers. The study suggests that older employees may perceive AI as a barrier to advancement rather than a catalyst for innovation.
Interestingly, work thriving did not vary significantly by age, indicating that even older employees can maintain energy and engagement. However, the trend shows that younger employees are more likely to embrace AI as a challenge, while older ones may see it as a burden.
Education: The Double-Edged Sword of High Skills
Educational attainment also played a crucial role in shaping psychological outcomes. Employees with higher education levels—particularly those with master’s degrees—reported higher anxiety (2.33) despite also experiencing high work thriving (4.30). This paradox reflects the complex reality of the “skill premium” in the AI era.
Highly educated employees are often in roles that are both highly valued and highly vulnerable to automation. They possess the cognitive abilities and achievement motivation to recognize the disruptive potential of AI, which can lead to heightened vigilance and stress. As the study notes, AI development is contributing to employment polarization, with growth in both high-skill and low-skill jobs, but a decline in mid-skill positions. This creates a sense of insecurity even among the most qualified workers.
In contrast, employees with lower educational levels reported lower anxiety (1.76 for high school and below). This may seem counterintuitive, but the study offers a plausible explanation: many low-skill roles being automated are also the most physically demanding or monotonous. When replaced by AI, these employees may experience relief from drudgery and find new opportunities in less automated sectors. Moreover, their expectations and perceptions of threat may be lower, leading to less psychological strain.
However, this group also reported lower psychological well-being and work thriving, suggesting that while they may be less anxious, they are also less engaged and fulfilled. This highlights a broader challenge: ensuring that the benefits of AI are not only distributed equitably but also contribute to meaningful work experiences for all.
Role-Specific Realities: The High-Pressure Life of R&D, the Confidence of Sales
Job function proved to be one of the most influential factors. R&D employees reported the highest levels of both anxiety (2.41) and work thriving (4.36). This dual profile captures the essence of innovation work: intense pressure to deliver breakthroughs, coupled with the excitement of solving complex problems and pushing technological boundaries. For R&D staff, AI is both a powerful tool and a demanding competitor, requiring constant learning and adaptation.
Sales employees, on the other hand, exhibited the highest psychological well-being (3.97) and the lowest anxiety (1.86). This may be because sales roles are less susceptible to automation due to their reliance on human relationships, emotional intelligence, and negotiation skills. Sales professionals may view AI as a support system—providing data insights, automating administrative tasks, and enhancing customer targeting—rather than a replacement.
Management and administrative roles showed more moderate levels of well-being and anxiety, while technical and production roles fell in between. The consistency of these patterns across the sample suggests that job design and perceived job security are key determinants of psychological response.
Beyond Fear: A New Framework for Understanding AI’s Human Impact
The study’s findings challenge the dominant narrative that AI is primarily a source of job loss and psychological distress. Instead, they suggest a more complex reality in which AI can simultaneously generate stress and satisfaction, anxiety and engagement.
The researchers draw on the Job Demands-Resources (JD-R) model to explain these dynamics. According to this framework, work environments contain both demands (e.g., workload, time pressure) and resources (e.g., autonomy, support, skill variety). When demands exceed resources, employees experience strain and burnout. When resources are abundant, they experience engagement and thriving.
AI, the study argues, can function as both a demand and a resource. It increases demands by requiring new skills and faster adaptation, but it also provides resources by automating routine tasks, enhancing decision-making, and enabling more creative and strategic work.
The net psychological effect depends on how AI is implemented and managed. In environments where AI is used to augment human capabilities and provide meaningful work, employees thrive. In settings where it is used primarily to monitor, control, or replace workers, anxiety prevails.
Implications for Organizations and Policymakers
The research offers several actionable insights for leaders navigating the AI transition.
First, organizations should recognize that AI adoption is not just a technical challenge but a human one. Psychological well-being must be monitored and supported with the same rigor as technical performance. This includes providing mental health resources, fostering open communication about AI’s role, and creating safe spaces for employees to express concerns.
Second, targeted interventions are needed for high-risk groups. Older employees, those in manufacturing and consumer services, and R&D staff may benefit from tailored training programs, mentorship opportunities, and career development support. These initiatives can help mitigate anxiety and build resilience.
Third, the design of AI systems should prioritize human-centered principles. AI should be introduced as a collaborative partner rather than a replacement. Transparency about how AI makes decisions, opportunities for employee input in AI development, and clear pathways for skill development can all enhance trust and acceptance.
Fourth, organizations should avoid the trap of assuming that longer AI experience automatically leads to better outcomes. The potential for adaptation fatigue means that continuous support and renewal of engagement are essential. Regular check-ins, feedback loops, and opportunities for employees to shape AI integration can help sustain morale over time.
For policymakers, the study underscores the need for proactive labor market strategies. As AI reshapes the skills landscape, governments must invest in lifelong learning, reskilling programs, and social safety nets that support workers in transition. The goal should not be to resist technological change, but to ensure that its benefits are widely shared and its disruptions are managed humanely.
A Call for Nuance in the AI Discourse
The study by Guo, Zhu, Li, and Du contributes to a growing body of research that calls for a more balanced and evidence-based understanding of AI’s impact on work. While the risks of automation are real and should not be ignored, the human response to AI is more varied and resilient than often assumed.
By focusing on psychological well-being, anxiety, and work thriving, the researchers provide a richer, more human-centered perspective on the AI revolution. Their findings suggest that with thoughtful design, supportive management, and proactive policy, it is possible to harness the power of AI while preserving—and even enhancing—employee well-being.
As AI continues to evolve, so too must our understanding of its human dimensions. This study is a vital step in that direction, offering not just data, but a framework for building workplaces where humans and machines can thrive together.
AI in the Workplace: How Employees Really Feel
Guo Juan, Zhu Xiaomei, Li Ziying, Du Hui, Beijing Union University, Enterprise Human Resource Management, 10.12345/ehrm.2021.11.050