Artificial intelligence has fueled fears about job losses for younger workers for years. New research now suggests the technology may also push many older professionals out of the workforce earlier than expected.
A June 30 paper from the Center for Retirement Research at Boston College found workers aged 55 and older in occupations with high exposure to artificial intelligence have become more likely to leave their jobs since the launch of ChatGPT.
Those departures include both voluntary exits and unemployment, raising new questions about how AI could reshape retirement, career planning and future Social Security reforms.
The research was written by Geoffrey Sanzenbacher, an economics professor at Boston College. It examined how generative AI has changed workforce behaviour among older employees.
Before OpenAI introduced ChatGPT, older workers in AI-exposed occupations tended to remain employed longer than peers in less exposed jobs. However, that pattern shifted after ChatGPT became widely available.
Sanzenbacher found older employees in AI-sensitive occupations now leave their jobs more frequently. Additionally, those departures appear evenly divided between voluntary exits and involuntary unemployment.
He said the effect is statistically significant and can be substantial in certain occupations.
The research outlines three possible ways artificial intelligence may affect career length for older workers. First, automation may replace workers entirely, leading to unemployment or permanent exits from the labour force.
Second, workers may choose different jobs that require less interaction with AI. Consequently, some may decide retirement offers a better alternative than adapting to rapidly changing technology.
The third possibility presents a more optimistic outcome. AI could increase productivity, raise wages and remove repetitive work, allowing employees to stay in the workforce longer.
Read more: Securitize begins NYSE trading with USD$266M tokenized stock launch
Read more: Coinbase wins UK approval for stock trading as it expands beyond crypto
Physical labour jobs among the least exposed
The study measured AI exposure by estimating how easily artificial intelligence can perform the tasks associated with specific occupations. Researchers combined data from the Current Population Survey with AI exposure scores developed through Tufts University’s Digital Planet initiative.
Those scores ranked web and technology professions among the most vulnerable to AI-driven disruption.
The occupations with the highest AI exposure include web and digital interface designers, web developers, database architects, computer programmers and data scientists.
Meanwhile, jobs involving physical labour ranked among the least exposed.
Excavating and loading operators, roof bolters and mining workers, orderlies, painting and spraying workers, and fiberglass laminators and fabricators recorded the lowest exposure scores.
Those findings challenge long-held assumptions about career length.
Traditionally, physically demanding jobs have been associated with earlier retirement because of their physical toll. However, Sanzenbacher’s findings suggest AI could narrow that gap by shortening careers in many white-collar professions instead.
The study found workers with high AI exposure also share several demographic characteristics. They are more likely to be white, hold college degrees and earn higher incomes than workers in occupations with low AI exposure.
That shift carries broader implications for retirement policy.
Social Security’s latest trustees’ report projects the trust fund supporting retirement benefits could become depleted in late 2032 unless lawmakers act.
Congress has several options to restore long-term solvency. Additionally, policymakers have repeatedly debated increasing the full retirement age.
The last major Social Security reforms occurred in 1983. Those changes gradually increased the full retirement age from 65 to 67.
Some experts believe lawmakers could revisit that approach during future negotiations.
Read more: MARA acquires Texas power site to expand AI and Bitcoin infrastructure
Read more: South Carolina signs crypto bill protecting Bitcoin mining and self-custody
Some older workers possess advances AI can’t replace
Sanzenbacher argued policymakers should consider AI’s effect on older workers before raising the retirement age further.
He also suggested higher-income workers could experience larger benefit reductions than lower-income retirees under future Social Security reforms. Those workers may need to remain employed longer to offset smaller retirement benefits. However, AI could simultaneously make remaining employed more difficult for some professionals.
That combination creates a difficult policy challenge.
Many higher-income workers occupy office-based jobs where AI adoption has accelerated rapidly. Consequently, the workers who may need longer careers could also face the greatest pressure from workplace automation.
Despite those concerns, researchers say older workers still possess advantages that AI cannot easily replace. Separate research conducted by AARP and LinkedIn found experienced professionals often work in positions less vulnerable to generative AI disruption than younger employees.
The analysis estimated 49.4 per cent of older workers hold relatively protected positions. Meanwhile, the figure for younger workers was 42.2 per cent. Researchers attributed much of that advantage to experience.
Older employees often perform leadership, collaboration and decision-making tasks that remain difficult for AI systems to replicate effectively. Those findings suggest technical knowledge alone will not determine future job security.
Instead, employers may place increasing value on communication, judgement, relationship management and problem-solving. Even so, many older workers remain cautious about AI adoption.
Sanzenbacher said older professionals generally use AI less frequently than younger colleagues. Recent AARP survey results illustrate that mixed outlook.
Among 1,015 adults aged 50 and older surveyed in March, 24 per cent viewed AI primarily as a threat to their work. Additionally, 19 per cent described AI as an opportunity.
Read more: ‘Largest AI cheating scandal in Ivy League history’: Brown professor appalled
Read more: Number of students using AI to cheat grows at alarming rate
Workers should become familiar with AI systems used by employers
Another 37 per cent said AI represented both a threat and an opportunity simultaneously. Career experts argue older workers still have time to adapt. Vicki Salemi, a career expert at Monster, said many employees have not yet embraced AI tools.
Monster’s December WorkWatch survey found 42 per cent of 1,504 workers surveyed do not use AI at all. Among workers already using AI, many rely on it for routine administrative work.
Common uses include writing emails, scheduling meetings and drafting documents. Others have adopted AI for more advanced work such as coding, automation, data analysis, résumé preparation and creative projects.
Salemi encouraged older professionals to pursue a two-part strategy. First, workers should become familiar with AI systems already used by their employers.
Additionally, learning those tools may free more time for strategic thinking and complex problem-solving. She also advised workers to strengthen the interpersonal skills AI cannot easily duplicate.
Communication, leadership, teamwork and relationship-building continue to carry significant value across industries. Those abilities may become even more important as automation expands into additional occupations.
Employers increasingly seek candidates who combine technical adaptability with strong interpersonal skills. Salemi said workers who demonstrate both qualities may improve their long-term employment prospects.
The Boston College research suggests AI’s labour market effects extend beyond entry-level employees and recent graduates.
Instead, the technology increasingly influences workers approaching retirement, creating new uncertainty for professionals who once expected longer careers in knowledge-based occupations.
As AI adoption continues across industries, researchers say understanding how technology affects different generations of workers will become increasingly important for employers, policymakers and retirement planners.
.
