Almost half of top managers define the implementation of artificial intelligence in their companies as a "huge disappointment" - against the backdrop of growing pressure on them to confidently lead their organizations through the AI transformation era. This conclusion is reached by a large-scale study published on Tuesday by the company "Writer" together with the independent research firm "Workplace Intelligence".
The basis of the report is a survey of 2,400 employees engaged in mental work, from the USA, Great Britain, Ireland, the Benelux countries, France and Germany. The collected picture outlines organizations caught between two fires: investments in AI continue to grow, and real returns remain difficult to achieve. Seventy-nine percent of managers admit that they are facing serious difficulties in implementing AI – this is a double-digit increase compared to 2025 – and only 29% report a tangible return on investment in generative AI tools.
Fear at the top
The gap between ambitions and actual results is causing great anxiety at the highest levels of business. Seventy-three percent of chief executive officers admit that they experience stress or anxiety in connection with their company's AI strategy, with 38% describing this tension as "high or unbearable". Almost two-thirds – 64% – share that they are afraid of losing their job if they fail to lead the organization through the transition to AI.
This fear leads to a phenomenon that the authors of the report call a "show strategy". Three-quarters of the managers admit that their company's AI plan exists "more for show" than as a real internal management tool, and 39% report that they do not have a formal plan to monetize AI tools at all.
"Mass layoffs are not an AI strategy," says Mei Habib, CEO and co-founder of Writer. "It is those leaders who are actually putting effort into radically readjusting operations with a focus on cooperation between people and agents who are increasing their lead so that competitors simply cannot catch up."
Two-tier workforce
Instead of investing in broad-based upskilling, many companies are betting on their most advanced AI users. Ninety-two percent of top managers say that they are purposefully forming an "AI elite" - employees who, according to the study data, are five times more productive than the rest and three times more likely to have received a promotion and salary increase in the last year. At the same time, 60% of managers plan to fire employees refusing to master AI, and 77% say that such workers will be bypassed in promotions.
This gap is generating open resistance. Twenty-nine percent of rank-and-file employees - and 44% of Generation Z representatives - admit that they have sabotaged corporate AI initiatives: they have entered company data into public tools, used unapproved applications or simply refused to work with the technology. Seventy-six percent of managers define this sabotage as a serious threat to the future of their companies.
Management problems exacerbate the crisis
Security concerns further exacerbate the situation. Two-thirds of senior managers believe that their company has already faced a data leak or breach as a result of using unauthorized AI tools by employees, and more than one-third admit that they would not be able to immediately disable an "out of control" AI agent that is causing damage.
"This is a turning point in the implementation of AI, and the gap between advanced users and those who are lagging behind is growing rapidly," comments Den Shobel, managing partner at Workplace Intelligence. Given that 97% of managers report the implementation of AI agents in the last year, but the return at the organizational level remains significantly lower than the growth of individual productivity, the key question for corporate leaders in 2026 is no longer "does AI work?", but "are their organizations able to build systems that turn this technology into real benefit?".