AI is rapidly scaling in the workforce and creating fears of an employment crisis, as workers and people entering the workforce try to figure out if their career is on the chopping block.
That quick pace is backed by emerging data. As a result, people are trying to find “AI-proof” jobs that can guarantee job security as companies around the world choose to automate tasks instead of hiring new workers.
Although no study can definitively say which occupations are 100% AI-proof and which are doomed to automation, a recent Microsoft study and its findings can shed a light on the matter.
A Microsoft study published last month measured how AI can productively apply to the common tasks of different jobs.
Microsoft researchers analyzed more than 200 thousand anonymized conversations from Bing Copilot, the company’s search engine chatbot, from January 2024 through September 2024 to see “what tasks users perform with a mainstream, publicly available, free-to-use generative AI chatbot,” the study says.
The study then developed “AI applicability scores” for these jobs, a number that represents the combination of which work activities people sought the most AI assistance for plus how successful these tasks were and their scope of impact.
There are caveats
Although the study shows which occupations AI can automate best, and those which it can’t do as well, Microsoft says that doesn’t necessarily mean that those jobs will be eliminated.
The AI applicability score highlights “where AI might change how work is done, not take away or replace jobs,” Microsoft representatives told Gizmodo earlier this month.
“Our research shows that AI supports many tasks, particularly those involving research, writing, and communication, but does not indicate it can fully perform any single occupation,” Microsoft said.
The data also does not imply that jobs with high AI applicability scores will have higher wages thanks to AI incorporation, the study noted, because the data does not include “the downstream business impacts of new technology.”
Read more about AI’s predicted effect on the corporate world from Gizmodo here.
Why companies automate
Microsoft believes AI can be used to augment these jobs rather than completely automating them.
But is that what corporate executives want? It’s tough to make a blanket statement on that, but early signs indicate that executives might be more pro-automation than not.
Increasingly, executives around the corporate world are voicing their expectations and desires to see AI cut costs across the workplace. This news has naturally led to a slowdown in hiring, particularly impacting early career workers in white-collar fields to which, as the Microsoft study also shows, AI poses the biggest threat.
“Artificial intelligence is going to replace literally half of all white-collar workers in the U.S.,” Ford CEO Jim Farley said at the Aspen Ideas Festival just last month.
Several executives have also already put into effect new hiring policies this year that ask managers to explain why an AI agent can’t fulfill the role before they can go ahead with hiring a new worker.
Just because you can doesn’t mean you should
AI can cut labor costs and increase profit for companies. But that is not yet a case for wholesale automation.
Although AI can automate some of these jobs, it doesn’t mean it can do a great job at it.
For example, Microsoft says that writers are in the top 10 for highest AI applicability. But AI-generated writing has been criticized far and wide, particularly for its bountiful copyright issues as AI feeds on the work of existing human writers to “create” new pieces.
The disruption of the labor market that is bound to follow the automation of certain jobs should also be a cause for concern.
Former Google executive Mo Gawdat said earlier this month that he believes this AI-driven labor problem is one of several aspects of the way we approach AI that is bound to lead to a short-term dystopia in the next 15 years.
Much like the Microsoft researchers that worked on the study, many other experts argue that the augmentation of AI into certain fields is a much better way to fuse AI into the economy for productivity gains than automation.
So what are the jobs?
Here are the ones most likely to stay human-run, the study says:
10. Tire Repairers and Changers
9. Ship Engineers
8. Automotive Glass Installers and Repairers
7. Oral and Maxillofacial Surgeons
6. Plant and System Operators
5. Embalmers
4. Helpers-Painters, Plasterers
3. Hazardous Materials Removal Workers
2. Nursing Assistants
1. Phlebotomists (aka healthcare professionals trained to collect blood samples)
AI works with data. So it is not surprising that the list overwhelmingly includes healthcare industry jobs, and blue collar work, both of which require specialized physical expertise rather than clear-cut data synthesis.
In the healthcare industry specifically, AI adoption has also been particularly slow due to limited datasets. Only less than 10% of surgical data is publicly available due to strict regulations.
The jobs that are at highest risk
Microsoft also looked at jobs that it deemed had the highest levels of AI applicability. Those were, rather unsurprisingly, knowledge work occupations and sales roles, where AI is already being rapidly incorporated.
Here is the list of the top 10 jobs that have the highest levels of AI applicability:
10. Broadcast announcers and radio DJs
9. Ticket agents and travel clerks
8. Telephone operators
7. CNC tool programmers
6. Customer service representatives
5. Writers and authors
4. Sales representatives of services
3. Passenger attendants
2. Historians
1. Interpreters and translators