(TNND) — Some groups of workers might have a harder time landing on their feet if they lose their jobs to artificial intelligence.

Identifying the highest-risk occupations can help lawmakers, businesses and communities funnel their limited resources towards workers who could most benefit from additional support.

That’s why researchers from the Centre for the Governance of AI and Brookings Metro examined the intersection of AI exposure and worker adaptability.

“If you think that some share of this impact is going to be job displacement, which seems unavoidable to some extent, even if it's not going to be mass unemployment anytime particularly soon, then a core policy challenge here is trying to understand how can we make workers more resilient to that change,” said the lead author of the new report, Sam Manning of GovAI.

The researchers found occupations with the highest AI exposure and then determined which are best and worst equipped for adapting to AI-driven job loss.

For example, office clerks and financial analysts are equally exposed to AI, according to the breakdown. But financial analysts scored a whopping 99% for adaptive capacity, while office clerks scored just 22%.

The researchers found that about 37 million Americans have jobs that fall within the top quartile of AI exposure, and most, about 70%, should be able to manage job transitions fairly well if necessary.

But about 6 million workers face both high AI exposure and low adaptive capacity, they said.

Those highest-risk jobs tend to be clerical or administrative, most are held by women, and many are concentrated in smaller metropolitan areas, particularly university towns and midsized markets in the Mountain West and Midwest.

“Many of the people who will be most exposed are also some of the most well-equipped to roll with the punches, whereas there are others who are not really well-equipped to get the next job after something goes wrong,” said Mark Muro, a senior fellow at Brookings Metro who researches AI and the digital economy.

AI exposure is not by itself a negative or harmful thing, Manning said.

Workers with strong financial resources, transferable skills and deep professional networks should be better prepared to weather the storm.

Occupations with high AI exposure that also have the highest adaptive capacity include web and digital interface designers, marketing managers, producers and directors, financial and investment analysts, computer and information systems managers, computer network architects, other mathematical science occupations, web developers, other life scientists, other financial specialists, information security analysts, software quality assurance analysts and testers, computer and information research scientists, chemists and materials scientists, and public relations and fundraising managers.

Highly exposed jobs with the lowest adaptive capacity include door-to-door sales workers, news and street vendors, and related workers; court, municipal, and license clerks; secretaries and administrative assistants (except legal, medical, and executive); payroll and timekeeping clerks; property appraisers and assessors; tax examiners and collectors, and revenue agents; eligibility interviewers for government programs; general office clerks; medical secretaries and administrative assistants; insurance sales agents; interpreters and translators; receptionists and information clerks; insurance claims and policy processing clerks; tax preparers; and legal secretaries and administrative assistants.

Muro said their research is “directional” but not predictive. Just because an occupation aligns with a lower capacity to adapt to AI disruptions doesn’t mean those workers are going to lose their jobs.

Other factors, such as location and the pace of AI adoption, matter, too.

RELATED STORY: Experts agree that AI is coming for jobs, but the switch might not happen quickly

Manning and Muro said an office clerk in a larger metro should have an easier time finding a new job if they’re displaced than an office clerk in a small town. There are simply more opportunities available in a bigger city.

And both men said a long, slow transition is best for workers to adapt. But that’s not guaranteed.

“Speed matters hugely here,” Muro said.

And he said there’s no time to delay for policymakers to get ahead of potential job disruptions.

“Because this may be something that happens slowly for a while and then happens suddenly at once,” he said.

Muro said they found relatively broad resilience to AI disruption among workers, but the research found significant pockets of concern.

Proactive action is better than trying to play catch-up, he said.

“A complete laissez-faire approach to this might well be a recipe for dissatisfaction and agitation,” Muro said.

SEE ALSO: Experts warn workers: stay tech-savvy as AI reshapes jobs in the next decade

Manning said there are no “super clean right answers” with policy to blunt the harms to workers, and he said a flexible and worker-specific approach would be best. Some workers might need new skills. Others might need help relocating.

“Workers are very diverse, and even within our occupation measures, there's a wide range of kind of adaptive capacity within a given occupation,” he said.

Manning said the government should shoulder most of the burden of balancing the benefits of AI with ensuring that workers can emerge on the other side with the ability to earn an income and support their families.