Artificial intelligence has quickly become one of the most widely used work tools among professionals worldwide. With its support, an increasing number of tasks can now be completed faster and more easily. This often creates the impression that so-called “AI-driven” workplaces are becoming more productive and innovative. In practice, however, the scale of AI adoption does not always translate into a genuine change in the way people work or into higher efficiency.
- 54% of specialists and managers use AI tools in their daily work, most often for simple and less demanding tasks.
- Professionals most commonly use AI for data analysis and research. It is used least often for generating creative ideas and writing or debugging IT code.
- The largest group of respondents, 55%, believe that AI-generated work requires only moderate refinement. This suggests that employees have a high level of trust in artificial intelligence.
AI technology is now ubiquitous in the labour market. However, the way it is used remains highly varied, and the level of maturity among organisations and employees does not always keep pace with the speed of technological implementation.
“Today, many employees and companies lack a conscious and sensible approach to using AI at work. Excessive focus on using it simply for the sake of using it, combined with limited understanding of real needs and possibilities, means that artificial intelligence can sometimes create more work for professionals rather than less,” says Alicja Malok, Senior Director at Hays Poland.
AI as a common workplace tool
Employees are experimenting with AI and increasingly recognise that many tasks, especially simple or repetitive ones, can easily be delegated to artificial intelligence tools. As a result, the use of AI at work continues to grow. This is confirmed by the latest Hays Poland survey, conducted in June 2026 among more than 700 professionals.
As many as 54% of specialists and managers say they use artificial intelligence at work every day, while another 24% use it several times a week. This shows that AI has quickly ceased to be viewed as a technological novelty and has become a basic work tool across many industries and job levels.
How often do you use AI tools at work?
| Frequency | Share of respondents |
|---|---|
| Every day | 54% |
| Several times a week | 24% |
| Several times a month | 8% |
| Less than once a month | 3% |
| I do not use AI at all | 11% |
Source: Hays Poland survey, June 2026.
At the same time, around one in ten respondents said they do not use artificial intelligence at work at all. This may be because the roles held by this group are not and will never be exposed to AI, or because their employers have not yet purchased business licences for employees.
However, this does not necessarily mean that professionals are not using the technology independently, outside the AI tools officially provided by their employers, for work-related purposes.
High adoption, but low maturity
The high level of AI use at work raises questions about the scope and effectiveness of this technology. Does artificial intelligence genuinely support employee efficiency, or does it mainly serve an auxiliary or experimental function?
The Hays survey shows that the tasks delegated to AI are predominantly relatively simple: data analysis, searching for and organising information, as well as writing or paraphrasing text. Professionals are much less likely to use AI for advanced and strategic projects that require technical skills, such as creating visual content or writing and debugging code, or creativity, such as generating ideas.
What do you most often use AI for at work?
| Use case | Share of respondents |
|---|---|
| Data analysis | 28% |
| Research and organising information | 24% |
| Support with writing or paraphrasing texts | 21% |
| Summarising texts | 17% |
| Decision-making support | 16% |
| Experimenting out of curiosity | 15% |
| Automating administrative tasks | 13% |
| Defining concepts or issues | 12% |
| Creating visual content | 11% |
| Generating creative ideas | 11% |
| Writing or debugging code | 7% |
| Other | 1% |
Respondents could select more than one answer.
Source: Hays Poland survey, June 2026.
This reflects a broader picture of the market.
“Many organisations are currently at an early stage of AI maturity. This means that artificial intelligence tools are present in companies, but they operate alongside existing processes rather than automating or transforming them. Today, the phrase ‘AI-driven’ in business is often simply an attractive buzzword. The mere presence of a tool does not yet mean a real change in the way work is done,” notes Alicja Malok of Hays.
In this context, there is also a risk of a phenomenon known globally as “tokenmaxxing”. It refers to situations in which employees, aware of growing pressure to use AI, do indeed work with artificial intelligence tools more frequently. However, this use may be more performative or declarative in nature, while the outcomes of AI collaboration do not always create added value for the company.
In practice, AI-generated results can be superficial, poorly adapted to the context or require further specialist work. Such output is sometimes referred to as “workslop”.
Too much trust in AI?
The fact that AI does not always make work easier or faster is also reflected in respondents’ answers about the quality of results generated by the technology. Fourteen per cent of respondents say that AI-generated content or recommendations usually require substantial corrections, while 9% believe it would be easier to complete the task independently.
This means that the burden is shifting from creating content to verifying it, editing it and taking responsibility for the final result.
At the same time, the survey found that a large group of professionals take a relatively optimistic view of AI’s reliability and quality. The largest group, 55%, believe that AI-generated work requires only moderate refinement, while 19% say it needs only minor changes.
This may partly result from the fact that employees mainly use artificial intelligence for simple tasks. Nevertheless, it still raises questions about trust in the technology, as well as the level of critical thinking and oversight applied to AI-generated output.
To what extent do AI-generated results, such as content or recommendations, usually require additional work before you can use them?
| Level of additional work required | Share of respondents |
|---|---|
| They require a lot of work; it would be easier to do it myself | 9% |
| They require major corrections | 14% |
| They require moderate refinement | 55% |
| They require minor changes | 19% |
| They are ready to use | 3% |
Source: Hays Poland survey, June 2026.
“In practice, many professionals still do not have sufficient knowledge of artificial intelligence and how it works internally. Even when employers provide AI tools, employees are often inadequately trained and unaware of the opportunities and risks associated with the technology. Using AI requires appropriate skills, both in formulating instructions and verifying results, especially as many employers expect employees supported by AI to become more efficient,” adds the Hays expert.
This does not mean that AI should not be used. On the contrary, by clearly defining current needs related to the use of this technology and then providing appropriate practical training, employees and companies can genuinely benefit from the potential of artificial intelligence.
The recommendation for both sides is therefore to carefully assess expectations and the scope of AI use at work, continue education and monitor trends, while adopting a gradual and practical approach to implementing artificial intelligence in everyday operations.
