News
Factorial Report: 78% of Employees Use AI at Work, but Only 43% of Companies Trained Their Whole Staff

A new Factorial report shows that nearly 80 percent of Polish employees already use AI in their daily work, but only 43 percent of companies have trained their entire staff on these tools. The authors warn of a growing gap between the pace of AI adoption and organizations' readiness to oversee it.
Contents
Polish companies are rolling out artificial intelligence faster than they are teaching employees how to use it. A new report from the European HR platform Factorial finds that 78 percent of employees already use AI in their daily tasks, yet only 43 percent of companies have trained their entire staff on these tools.
The report, titled "The Hidden Power of Artificial Intelligence," is based on a survey conducted by Factorial, a European AI-powered human resources management platform. The authors compared employees' own accounts with data on how organizations actually manage the rollout of new tools.
Adoption Outpaces Oversight
The report's key finding concerns speed. Over three years, the share of companies that have deployed or are deploying AI rose from 62 to 77 percent. At the same time, employee use of these tools has grown just as fast, with people turning to chatbots and AI assistants regardless of whether their employer has formally approved the practice or trained them to use it safely.
Only 55 percent of companies have a written, formal AI strategy, and 29 percent are still working on one. The picture is even worse when it comes to rolling out concrete usage rules: 71 percent of organizations have started the process, but 16 percent have stalled, and 8 percent have yet to even begin.
Employees Fear Losing Control, Not AI
The survey shows employees are not rejecting the new tools. As many as 78 percent say they are willing to let AI take part in decision-making, though 48 percent insist on being able to verify the results, and 30 percent limit that trust to repetitive tasks only. At the same time, only 42 percent of respondents say they understand how the models they use every day actually work.
We are buying innovation faster than we are learning to manage it - Tomasz Wykowski, Country Manager Poland, Factorial
Wykowski points out that the problem isn't enthusiasm for AI itself, but the lack of procedures that keep pace with it. In another comment cited in the report, he adds that employees' unease is less about the technology itself than about losing control over their own work.
Employees aren't afraid of artificial intelligence itself, they're afraid of losing control - Tomasz Wykowski, Country Manager Poland, Factorial
Data Security in Question
The report also points to a gap in cybersecurity. While 96 percent of companies say they have carried out some form of AI risk assessment, more than a third admit that real control over what data reaches AI tools remains limited. This mirrors what other studies describe as "shadow AI", employees using personal accounts and AI apps without the IT department's knowledge, often to process company or client data.
A similar picture emerges from Microsoft and LinkedIn's Work Trend Index, according to which 75 percent of office workers worldwide use AI at work, and 78 percent of them do so using their own, unapproved tools. The scale of the phenomenon is also growing within companies' own network-monitoring systems, where the number of queries sent to generative AI is climbing several-fold month over month.
What It Means for Polish Companies
For Polish employers, the Factorial report is a signal that simply buying licenses for AI tools is not enough. Without organization-wide training and clear data rules, companies risk a situation where the technology is already part of everyday work, but no one is formally accountable for how it's used.
The report's authors recommend that companies treat AI training not as a one-time rollout, but as an ongoing part of skills development, covering both how to use the tools and awareness of the risks tied to data security and accountability for decisions made with the help of algorithms.

