Artificial intelligence may be shortening careers for some older workers, according to new research that finds employees in AI-exposed occupations have become more likely to leave work since the launch of ChatGPT.
Researchers compared employment patterns before and after ChatGPT’s release in late 2022 to examine whether workers aged 55 and older in AI-exposed jobs were leaving the workforce at higher rates, according to The Center for Retirement Research at Boston College.
Higher AI Exposure Linked to More Job Exits
Before ChatGPT, workers in occupations with higher AI exposure generally stayed employed longer than those in less-exposed roles, largely because those jobs tend to be higher paying, less physically demanding, and require more education.
That advantage has weakened since generative AI entered the workplace.
Researchers found that older employees in occupations with greater AI exposure are now more likely to leave work, with much of the increase driven by transitions into unemployment rather than retirement. The findings suggest some late-career workers may be struggling to remain employed as AI changes job requirements.
Knowledge Jobs Face the Greatest Risk
The study measures AI exposure based on how much of a job’s work can potentially be performed by AI, rather than whether the occupation is likely to disappear entirely.
Occupations centered on data analysis, software development, and other knowledge work ranked among the most exposed, while hands-on jobs requiring physical labor or direct interaction with people ranked among the least exposed.
Researchers estimate that highly exposed occupations could experience substantially larger increases in job exits than lower-exposure roles, although workers in those occupations still tend to have longer careers overall.
While AI may improve productivity for many employees, the report suggests workers nearing retirement may face additional pressure if they struggle to adapt to rapidly changing technology or lose jobs in AI-exposed occupations.
