The debate about the impact of artificial intelligence on the labor market is currently dominated by one key question: will AI lead to the mass disappearance of jobs? However, the latest empirical research paints a much more complex picture.

Raport The report “Labor Market Impacts of AI: A New Measure and Early Evidence” (2026) by Maxim Massenkoff and Peter McCrory analyzes the real-world use of generative AI at work and its potential impact on employment. The authors introduce a new approach to measuring how exposed different occupations are to AI by combining the theoretical capabilities of language models with real data on how they are actually used in the workplace.

The report’s findings are not only relevant for the U.S. labor market, which was the subject of the study. They also point to trends that are highly likely to emerge in Europe, including Poland.

The gap between AI’s potential and its actual use

One of the report’s key findings is the clear gap between the technological capabilities of AI and its real-world application at work.

In theory, language models could accelerate a large share of tasks across many professions. In practice, however, AI is currently used for only a fraction of these possibilities. For example, in IT and data analysis roles, AI could potentially support most tasks, yet its actual use covers a much smaller portion of daily work.

This gap results from several factors. On one hand, there are regulatory constraints and the need for human quality control. On the other, there is a lack of full integration of AI tools into organizational systems, as well as the natural inertia of companies when it comes to adopting new technologies. As a result, technology is advancing faster than its implementation.

For companies, this means one thing: the biggest changes are still ahead, and the pace of transformation will depend more on organizational adaptability than on technological progress itself.

AI automates tasks, not entire jobs

The report also confirms an important pattern: AI rarely replaces entire occupations. Instead, it tends to take over individual tasks within them.

Each job consists of multiple activities that vary in complexity and susceptibility to automation. Generative AI performs best in tasks involving information processing, data analysis, and content creation. As a result, some operational work can be automated, while other aspects of a role remain human-driven.

This leads to a gradual shift in the nature of work. Routine tasks are becoming less important, while interpretation, decision-making, and the ability to connect knowledge across domains are gaining value.

Information-based roles are the most exposed

The report clearly shows that occupations based on working with information are the most exposed to AI.
These include roles centered on data analysis, text processing, documentation, or coding—such as positions in IT, business analytics, finance, customer service, and marketing. These are precisely the areas where AI is already widely applied and rapidly expanding.

This represents a reversal of earlier waves of automation, which primarily affected physical labor. This time, the transformation is focused on white-collar and analytical work—areas previously considered less vulnerable to automation.

Does AI increase unemployment?

One of the most important conclusions of the report is the lack of clear evidence that AI is increasing unemployment in highly exposed occupations. Analysis of U.S. labor market data shows that, following the spread of generative AI, unemployment levels in these groups have remained stable.

This does not mean there are no changes. Instead of mass layoffs, we observe more subtle shifts, particularly in the structure of employment. Notably, there are signs of slower hiring among young workers in occupations most exposed to AI.
This may be because tasks typically performed by entry-level employees are now being supported or partially replaced by AI tools. As a result, the way people enter the labor market is evolving.

What does this mean for Poland?

Although the report is based on U.S. data, its conclusions are largely universal and directly relevant to Poland.

In recent years, Poland’s economy has strongly developed its business services sector, including shared service centers, analytics, financial processes, and IT support. These are exactly the areas identified as most exposed to AI.
This means that the impact of generative AI in Poland may be particularly visible in these segments of the labor market. At the same time, a simple scenario of job losses is unlikely. A more probable outcome is the transformation of roles—reducing operational work while increasing the importance of analytical and decision-making skills.

This shift may also affect business models based on cost advantages, which have long underpinned the growth of Poland’s service sector. As automation increases, expert capabilities and the ability to create added value will become more important.

The biggest change: transformation, not disappearance

The most important takeaway from the report is somewhat paradoxical. Contrary to widespread fears, the biggest change may not be the mass disappearance of jobs, but rather how they are performed.
AI is taking over some operational tasks, leading to a gradual transformation of professional roles. In many cases, human work is shifting toward supervision, interpretation, and decision-making.
Organizations will need to rethink how teams operate and redefine roles, while employees will face a growing need for analytical, interdisciplinary skills and the ability to collaborate with technology.

The impact of AI on the labor market is not a sudden revolution. It is an evolutionary process, the pace of which will depend primarily on how quickly organizations learn to harness the potential of this technology.