Jennie Brand, UCLA professor of sociology with an appointment in statistics and data science, is an expert on the socioeconomic effects of job loss, including the long-term economic, psychological, physiological and intergenerational effects.
As the current job landscape is being transformed by artificial intelligence, Brand and her research partners at the Occupation Lab have been examining how occupational change, including AI-exposure, is affecting workers’ job mobility. An AI-exposed job is one in which tasks could potentially be performed more efficiently with AI.
In their recent paper published in Science Advances, “Trapped in declining occupations: Barriers to worker mobility in a changing economy,” Brand and her research team found that workers employed in growing occupations have better prospects for upward mobility than those employed in declining occupations — creating a “mobility trap” for the latter group. Findings specific to AI-exposed jobs were also recently published in the Papers and Proceedings of the American Economic Association.
Brand, who is also co-director of the UCLA Center for Social Statistics, and Xi Song, who received her Ph.D. from UCLA and is now a professor at Columbia University, answered questions related to their research.
What is an AI-exposed job?
An AI-exposed job is one whose job tasks could potentially be performed more efficiently with AI. In our research, we consider a job AI-exposed if large language models and related software could reduce the time needed for particular tasks by at least half while maintaining quality. The number of people in an occupation highly exposed to AI might decline, grow or change substantially as workers use AI. We also distinguish AI exposure from AI skills: having a job that AI can affect does not necessarily mean having the skills or support to work effectively with it.
Who did you find is most at risk of losing their job to AI — and why?
We highlight a particularly concerning situation: Workers in occupations that are both highly exposed to AI and projected to decline in terms of the number of people in those occupations. These workers face pressure on their existing jobs and may face limited mobility paths into expanding fields. Our work finds that changing occupations often results in moving into another declining occupation without improving their economic position, a “mobility trap.” The central concern is therefore the combination of AI exposure, declining demand and limited access to better alternatives. Exposure alone is insufficient to identify who is most vulnerable.
Which specific jobs and industries are most impacted by AI?
Our preliminary research on AI skills identifies high AI exposure in occupations such as payroll clerks, insurance claims and policy-processing clerks, bookkeeping clerks, medical transcriptionists, data-entry workers, and interpreters and translators. These occupations involve information processing, documentation or language tasks. They occur across insurance, healthcare, accounting, administrative services, and other industries. Some scientific and technical occupations are also highly exposed, but workers in these fields report more AI-related skills (e.g., machine learning researchers). AI can therefore create very different challenges and opportunities across occupations.
Which groups are most vulnerable to AI affecting their job prospects and status?
Our preliminary findings highlight a potentially vulnerable group: workers in occupations with high AI exposure but whose profiles show little evidence of AI-related skills. For example, less than 1% of payroll and bookkeeping clerks in our data from LinkedIn list a generative AI skill on their profiles. Vulnerability may also be greater for people who lack time, money, employer training or access to alternative jobs, although our current research does not directly test all these mechanisms.
What does job churning mean, and why should a layperson care?
Job churning refers to a pattern in which workers repeatedly leave, enter or change jobs. Some movements reflect opportunity, while others reflect instability. Our research emphasizes that movement itself does not tell us whether people are getting ahead. Someone can change occupations yet remain in a declining field or move into lower-paid work. Conversely, little movement can reflect either security or difficulty finding alternatives. A key feature is whether a move improves pay, security, learning and long-term career development.
How does losing a job and having to change jobs affect well-being?
Losing a job can disrupt income, daily routines, social relationships and a person’s sense of identity. Finding another job may not resolve all the consequences of displacement. While a voluntary move to better work can be beneficial, an involuntary move with weaker prospects is a different experience. In our recent work, we are considering workers displaced from AI-exposed occupations and how their career trajectories differ across selected characteristics. Workers who lose jobs respond differently to those experiences based on their education, skills, age and other characteristics, including AI skills. We are using machine learning models to identify sources of variation in displaced workers’ transitions.
How can I use AI so I don’t get left behind in the job market?
AI tools are evolving quickly and becoming easier to use, even without programming skills. Learning to use these tools can be valuable, but field expertise remains necessary to judge whether AI’s output is accurate and useful, and to improve it when it contains errors. AI can help workers work more efficiently, but over-relying on it can undermine the learning and practice that build expertise.
Will AI lower my leverage to ask for a raise or promotion?
It could, but the effect is not automatic. If employers can readily replace some of a worker’s tasks and that worker has few alternative job opportunities, bargaining power may weaken. If AI helps workers deliver valuable results while expertise remains difficult to replace, a worker’s bargaining power can even improve. Our research highlights the importance of outside options: workers in declining occupations can struggle to reach growing fields.
What should employers and policymakers do so that AI-related disruption does not trap workers in worse careers?
Employers and policymakers could focus on alleviating the difficulties of layoffs, and creating and strengthening stepping-stone jobs that provide pathways into growing occupations. Policies could connect job training to actual openings in growing fields, help workers demonstrate transferable skills and support them during transitions. Employers could provide paid learning time and internal routes into new roles. These are recommendations motivated by our findings, not interventions our papers have evaluated.