Artificial intelligence and the job market: which professions are most at risk and who can actually adapt
Artificial intelligence is reshaping the job market at a speed that caught a lot of people off guard. This is no longer some distant conversation about robots replacing humans in a far-off future. The shift is already happening, and some professions are feeling the weight of this transformation in their day-to-day work. At the same time, there are fields that remain virtually untouched by automation, and the divide between who is affected and who is not has some very concrete explanations. A study recently published by GovAI, an organization dedicated to technology policy research, in partnership with the Brookings Institution, one of the most respected think tanks in Washington, shed light on this landscape in a way that goes beyond the obvious. Instead of just listing which jobs are in AI’s crosshairs, the researchers investigated an aspect that usually gets left out of the conversation: the ability each worker has to reinvent themselves when change comes knocking.
The study analyzed thousands of occupations and cross-referenced data on automation exposure with indicators of individual vulnerability. This includes factors like a worker’s age, education level, available financial savings, and the possibility of transferring skills to other roles. The result is a remarkably detailed map that shows not only who is most exposed, but also who has more or fewer resources to bounce back.
What the GovAI and Brookings Institution study revealed
The researchers used a novel approach to estimate which workers would have a greater or lesser ability to adapt to artificial intelligence. The finding that grabbed the most attention is that many of the people at highest risk of having their jobs transformed by AI are also the best positioned to find new occupations. It sounds contradictory, but it makes perfect sense when you look at the data closely.
Web designers, for example, are among the professionals most exposed to automation. AI tools can already generate layouts, prototype interfaces, and even write front-end code at an impressive speed. Even so, most web designers should be able to land on their feet, precisely because they have advanced digital skills, creative thinking, and familiarity with emerging technologies — abilities that are highly transferable to other roles in the market. Secretaries and administrative assistants, on the other hand, face a very different scenario. Their tasks are also directly in AI’s sights, but these professionals tend to have a harder time transitioning, due to factors like lower salary ranges, less accumulated savings, and skill sets that don’t always migrate easily to other sectors.
One data point that deserves special attention is the gender breakdown. According to the study, the most vulnerable occupations — those that combine high AI exposure with low adaptability — are predominantly held by women. This means the impact of artificial intelligence on the job market is not socially neutral. It could deepen inequalities that already exist if no concrete steps are taken to mitigate the effect.
Who is most exposed and who can adapt
When we dig deeper into the data, it becomes clear that exposure to artificial intelligence does not hit every sector the same way. Professions involving repetitive tasks, information processing, and standardized data analysis are among the most impacted. Administrative roles, routine financial positions, and even some jobs in the legal sector show up with a high degree of exposure.
Activities that rely heavily on in-person human interaction, unpredictable creativity, or complex manual skills remain at a much lower risk level. Cleaning and building maintenance workers, electricians, nurses, and industrial maintenance staff, for instance, continue to sit in a relatively comfortable zone — at least for now. The study specifically notes that janitors and custodians face less risk than web designers and secretaries, which illustrates well how AI is extremely efficient at structured cognitive tasks but still stumbles quite a bit when it needs to deal with unpredictable physical environments or situations that require empathy and refined human judgment.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
But the real insight of the study lies precisely in the distinction between exposure and vulnerability. A professional can be highly exposed to automation and still have good chances of finding a new path. This is the case, for example, with data analysts and software developers who possess transferable skills for areas like project management, strategic consulting, or developing AI-based solutions. These workers typically have solid educational backgrounds, access to diverse professional networks, and financial reserves that allow them to invest in reskilling without panic.
On the flip side, there are professionals in roles less exposed to AI who, if caught off guard by automation, would face a much tougher situation. Older workers with less education and little savings make up the group that researchers classified as the most vulnerable — not necessarily because their jobs are disappearing right now, but because they would have very few resources to reinvent themselves if it happened.
Why predictions about AI and employment are so contradictory
If you follow the news about artificial intelligence and the job market, you have probably noticed that predictions contradict each other all the time. And it is not just your imagination. The experts themselves acknowledge that we are navigating uncharted waters.
An influential analysis from Stanford University published last year concluded that AI is likely already eliminating positions among young professionals in fields like software development and customer service, where the adoption of artificial intelligence tools has been the fastest. However, a different study conducted by the think tank Economic Innovation Group reached practically the opposite conclusion: young workers in those same AI-exposed occupations were actually doing better than peers in less affected fields like personal training and construction.
To make things even murkier, the Federal Reserve Bank of Dallas stated that it is unlikely AI will eliminate jobs in any significant way over the next decade. Meanwhile, CEOs of major tech companies keep predicting that millions of people will lose their positions soon. This rollercoaster of forecasts can make anyone dizzy 😵💫
Jed Kolko, a senior fellow at the Peterson Institute for International Economics, summed up the situation well by saying that all the important questions about the effects of AI on the labor market remain unanswered. Economists at Anthropic, the AI startup behind the chatbot Claude, were even more direct and stressed the need for humility when analyzing how artificial intelligence is penetrating different occupations. According to them, recognizing the limits of what we know is essential to avoid jumping to conclusions — a posture that is not exactly common in Silicon Valley, to be fair.
History does not help much either for anyone looking for certainties. Economists and researchers have a pretty rough track record when it comes to predicting the effects of new technologies on work. Previous technological revolutions — from the steam engine to the personal computer — always generated catastrophic predictions that never played out as expected. That does not mean this time will be the same, but it serves as a reminder to take any projection seriously without treating it as absolute truth.
The social dimension that cannot be ignored
One of the most important aspects the study brings to light is the social dimension of this transformation. We are not just talking about technology replacing tasks. We are talking about real people with stories, limitations, and very different life circumstances.
The research makes it clear that vulnerability to AI is directly tied to socioeconomic factors. Professionals with a bigger financial cushion can afford to take career breaks to study, invest in upskilling courses, and even accept temporarily lower-paying positions while making the transition to a new field. Those living paycheck to paycheck simply do not have that kind of flexibility. Any interruption in income can mean an immediate financial crisis.
Age also plays a crucial role. Younger workers tend to have an easier time learning new tools, switching fields, and adapting to new work environments. More experienced professionals, while bringing a valuable repertoire of accumulated knowledge, often face additional barriers — from age discrimination in the market to less familiarity with the latest technologies.
On top of that, the fact that the most vulnerable occupations are predominantly held by women adds an important layer of concern. If automation advances without adequate support policies in place, the result could be a widening of gender disparities in the job market, reversing gains that took decades to achieve.
What this means for the future of work
The findings presented by GovAI and the Brookings Institution point to a conclusion that should guide decisions by governments, companies, and workers alike. Artificial intelligence is not going to simply destroy jobs en masse — it is going to redraw the board unevenly. Some roles will disappear, others will transform, and many new ones will emerge. The problem is that this redistribution does not happen in a balanced way.
Those with more resources, more education, and more flexibility will navigate this transition with much greater ease. Those in tight financial situations, with little room to invest in training and without a strong support network, could end up being pushed into even more precarious positions. Individual adaptability is just as important as the level of technological exposure, and that is perhaps the study’s main takeaway.
This finding challenges a simplistic narrative that dominates much of the discussion around AI and work. The idea that all you need to do is learn to code or become a tech specialist to be safe is an incomplete picture of reality. Adaptation depends on a much broader set of conditions, and ignoring that can lead to ineffective policies that protect those who are already doing well while leaving behind those who need support the most.
Smart adaptation: what the data is telling us
For anyone following the tech market, it is becoming increasingly clear that the conversation about AI and work needs to mature. There is no point in staying at the extreme of alarmism, predicting a job apocalypse, or at the extreme of blind optimism, insisting everything will sort itself out. Reality is somewhere in the middle, and it demands concrete action.
Companies investing in internal reskilling programs are heading in the right direction. Governments creating safety nets for workers in transition are too. And professionals who are working to diversify their skill sets — learning not just technical tools but developing interpersonal skills, critical thinking, and the ability to solve complex problems — are positioning themselves more resiliently in the face of uncertainty.
The study reinforces that professional reskilling programs need to be designed with the full profile of the worker in mind, not just the degree of technological exposure their current role faces. Considering age, financial situation, transferable skills, and even geographic location can make the difference between a program that actually works and one that only looks good on paper.
Vulnerability to AI is not determined by technology alone, but by each person’s socioeconomic context. Ignoring that factor is a recipe for amplifying inequalities that already exist. At the end of the day, adaptation really is the keyword in this story. There is no magic formula that guarantees absolute security in the job market as artificial intelligence advances. What does exist is the chance to prepare more intelligently, understanding where the real risks are and, most importantly, recognizing that the ability to change course is an asset just as valuable as any degree or technical certification.
The data is right there, and it tells a story that is more complex and more human than most headlines can capture 🧠
