Artificial intelligence is generating a striking gap between what companies prepare and what their workers expect. While a significant portion of employers already contemplates smaller workforces thanks to automation, very few employees believe that their own position will completely disappear.
The 41% of employers surveyed by the World Economic Forum stated that they expect to reduce their workforce by 2030 where the new capabilities of artificial intelligence allow for the reproduction of tasks currently performed by people. This data comes from the Future of Jobs Report 2025, compiled from the responses of more than 1,000 large employers representing over 14 million workers in 55 economies.
The perspective from the other side of the table is very different. A study published this summer by Infosys Knowledge Institute, based on more than 2,600 office professionals, concludes that only 2% of employees expect their position to be eliminated due to artificial intelligence. Nearly 45% even declare themselves excited about its use at work, and around 40% express confidence, motivation, or optimism regarding these tools.
41% expect cuts, but 77% also want to train their workers
However, business plans do not point to a simple mass replacement of people by machines. The same study from the World Economic Forum that reports possible cuts shows that 77% of companies intend to train or upskill their workforce to work alongside AI, while 69% plan to hire specialists capable of designing or improving these tools and 62% seek professionals with sufficient knowledge to use them.
Furthermore, 47% of employers intend to transfer workers from positions affected by automation to other internal functions, which outlines a more complex scenario than that of a technology that simply eliminates jobs. AI can make certain tasks disappear while creating other needs within the same company.
The World Economic Forum estimates that the transformation of the labor market caused by technology, demographic changes, the energy transition, and other factors could lead to the creation of 170 million jobs by 2030 and the displacement of another 92 million, with a net positive balance of 78 million. This does not mean that those who lose a job will automatically take one of the new ones: the main difficulty will be precisely that the demanded skills will be different.
Workers do not believe that AI will completely replace them
The Infosys survey helps to understand the apparent calm of a good part of the employees. AI has already been incorporated into everyday work and about 70% of the professionals surveyed claim to use it at least once a day for activities such as conducting searches, summarizing documents, researching, or generating content.
Many workers are experiencing technology first as a productivity tool and not as a substitute. In sectors like retail, for example, a significant portion of those using AI claims to save several hours a week and then use that time to perform more tasks or work considered of higher value.
This may explain why only a small minority expects their position to completely disappear, although the calm is not absolute either. About a quarter of the employees analyzed by Infosys acknowledges some insecurity and fears that their functions may be reduced even after receiving training to adapt.
The risk may be more in tasks than in complete positions
One of the common mistakes when analyzing the labor impact of artificial intelligence is to pose the question solely as a choice between keeping or losing a job. In many cases, the transformation can occur within the job itself.
An administrative worker may still be needed even if AI writes emails, prepares reports, or processes documentation. A programmer may continue working even if part of the code is generated automatically. And a customer service employee may keep their position while automated systems handle the more repetitive inquiries.
In fact, the World Economic Forum estimates that 39% of current workers' skills will change or become obsolete between 2025 and 2030. Skills related to AI and big data, networks and cybersecurity, or technological literacy are among those that will grow the fastest, but important human skills such as analytical thinking, creativity, flexibility, and leadership will also remain significant.
Administrative jobs among the most exposed
Not all professionals face the same level of risk. Positions composed of repetitive tasks, easily standardizable and primarily developed in front of a computer, are among the most exposed to automation.
The World Economic Forum ranks among the occupations with the greatest expected decline cashiers, ticket sales employees, administrative assistants, executive secretaries, postal service workers, bank tellers, and data entry professionals. At the same time, occupations related to artificial intelligence, software development, data, and cybersecurity appear among those with the highest expected growth.
The transformation will also be uneven across sectors. In technology services, for example, 49% of companies expect to reduce staff whose skills lose relevance, eight points above the overall average, although those same companies expect to continue hiring specialized profiles.
Job losses directly linked to AI are still limited
Business forecasts should not be confused with what has already happened. Although companies like Cloudflare have announced significant layoffs this year coinciding with increased internal use of artificial intelligence tools, the aggregated evidence still does not show massive job destruction attributable solely to this technology.
Recent studies of companies in the United States, United Kingdom, Germany, and Australia conclude that around 70% of companies already use some form of artificial intelligence, but that so far its overall impact on employment and productivity remains relatively small.
This also forces a careful interpretation of the famous 41%. It does not mean that four out of ten companies will lay off 41% of their workers, nor does it even mean that all will ultimately implement cuts. It means that four out of ten employers are considering reducing staff in those functions where AI can replace human tasks within their plans until 2030.
A gap in perception between those who decide and those who execute
The difference between companies and workers can become one of the main problems of the transition. If employees see AI simply as a tool to work faster while management studies in parallel reorganizing teams and reducing certain positions, the lack of information can generate a strong trust problem.
Infosys precisely warns of this situation: although only 2% believe that their job will disappear, more than half of the workers show some degree of uncertainty about whether their position will really be protected as tools become more capable.
Training will be one of the main responses. The problem for companies is not only having artificial intelligence but finding workers capable of using it correctly. The lack of skills is already the main obstacle to its implementation, cited by half of the executives surveyed by the World Economic Forum, even ahead of technology costs or regulation.
Neither a massive disappearance of jobs nor a painless transition
The data thus paints a situation less straightforward than the two most common narratives about AI. There is not enough evidence to assert that it will cause inevitable mass unemployment, but it also does not seem realistic to assume that its incorporation will not destroy certain positions.
Companies are already anticipating cuts in some areas while simultaneously seeking to hire new capabilities and retrain their employees. Workers, meanwhile, primarily perceive the immediate advantages of using tools that reduce repetitive tasks and save time, which helps explain why so few believe that they will be the ones who end up losing their job.
The great unknown of the coming years will be precisely in that space between both perceptions. Four out of ten employers already contemplate smaller workforces where AI can do the job, but only 2% of employees expect to be completely replaced. Between both figures, a good part of the labor transformation that artificial intelligence will bring will be played out.