AI is changing the jobs of older workers, study finds

Artificial intelligence has raised concerns that workers – especially young professionals – could be replaced by advancing technology.

However, there are emerging signs that older workers are also being affected, according to a new paper from the Center for Retirement Research at Boston College.

Workers age 55-plus in industries exposed to AI are leaving their jobs more often, author Geoffrey Sanzenbacher, a professor of economics, told CNBC. Those changes are driven equally by unemployment and voluntary reasons, he said.

“It’s a statistically significant result,” Sanzenbacher said. “For some jobs, it can be really big.”

AI may affect the career length of older workers in three ways, according to Sanzenbacher’s research.

Self-employment may take over older workers, causing them to become unemployed or leave the workforce altogether, Sanzenbacher found.

Even the pressure to adopt AI may lead some workers to seek jobs that don’t involve that new development or to retire altogether.

Alternatively, productive AI could allow people to work longer hours as productivity increases, raising wages and allowing workers to focus on more engaging tasks, according to Sanzenbacher.

The study defines AI exposure based on the extent to which AI can be used to perform work tasks. It uses Current Population Survey data and information about exposure to AI from Tufts University’s Digital Planet program, which studies the impact of digital innovation.

Research has found that prior to the launch of OpenAI’s ChatGPT, older workers in AI-exposed roles they are less likely to leave their jobs. However, after the introduction of ChatGPT, they are at a great risk of being out of work, including unemployment.

Older workers most affected by the AI ​​revolution tend to be white, more likely to have a college degree and tend to have higher salaries compared to workers with lower AI exposure, according to Sanzenbacher’s research.

According to the study, based on Digital Planet’s AI exposure scores, the five occupations with the highest exposure to AI – and therefore the most vulnerable to change – are:

• Web and digital designers
• Web developers
• Web site builders
• Computer programmers
• Data scientists

The five jobs with the least exposure to AI are:

• Digging and loading operations and miners
• Roof bolters and miners
• They are not organized
• Painters and sprayers
• Fiberglass laminators and fabricators

This calls into question the idea that older workers in heavy occupations may have shorter career trajectories and therefore retire before more educated, white workers.

“AI exposure may reduce the career length gap between low-paying and high-paying jobs,” Sanzenbacher wrote in the study.

Therefore, as policymakers consider possible changes to the retirement age, they should keep in mind the potential effects of AI, according to the study.

New projections show the trust fund Social Security relies on to help pay retirement benefits could run out as early as 2032, according to the latest annual report from the program’s trustees.

To restore Social Security’s solvency, policymakers can choose from a variety of possible reforms, including raising the retirement age. When the last major changes to the system were made in 1983, lawmakers chose to gradually raise the retirement age from 65 to 67.

Raising the age limit may be considered in future reforms, although some experts argue that the change will not be a temporary solution. Policymakers could also choose to raise payroll taxes on high earners, a popular proposal among Democratic lawmakers.

“High-income earners are more likely to see larger benefit cuts than low-income earners in whatever happens with Social Security next,” Sanzenbacher said.

“These are the very people who have to work for a long time,” he said.

Meanwhile, depending on how AI’s effects on workers play out, those same workers’ ability to do their jobs could also be affected, Sanzenbacher said.

Older workers are embracing AI, but not as often as younger workers, Sanzenbacher said.

A recent AARP survey found that workers see AI as both a threat and an opportunity. While 24% of the 1,015 adults aged 50 and over surveyed said they see AI as a threat to their jobs, 19% said they see it as an opportunity. Meanwhile, 37% said both, according to responses filed in March.

A separate study from AARP and LinkedIn found that experienced professionals are likely to hold roles that are hindered by disruption caused by generative AI, with 49.4% of older workers compared to 42.2% of younger workers.

The jobs of older workers may require skills that AI cannot easily replicate, according to research, such as cooperation, judgment and leadership.

For older workers who haven’t started using AI, it’s not too late, Vicki Salemi, an operations specialist at Monster, told CNBC.com in an email.

Monster’s December WorkWatch report found that 42% of the 1,504 workers surveyed are not using AI at all. Among employees who use it, the most popular applications include basic tasks such as email, editing and writing support, Monster found. Others use it for advanced tasks, such as coding, automation or data analysis; job applications, including resumes and cover letters; or creative work, such as graphics, campaigns and content, according to the study.

Older workers may want to take a two-pronged approach to adapting to today’s workforce — becoming AI-literate while also doubling down on soft skills, Salemi said.

A good place to start would be to learn the AI ​​tools your employer is already using, which can help increase your time for critical thinking, he said.

At the same time, don’t forget to highlight soft skills such as communication, relationship building and problem solving, whether you want to stand out in your current role or the job you’re applying for.

“If you can demonstrate that you have strong soft skills combined with the ability to evolve and grow with new technologies, it can be a green light for your selection,” said Salemi.