News
Nearly Half of Polish Companies Disappointed with AI Rollouts, Yet Spending Set to Rise
A new EY report finds that 49 percent of Polish companies are disappointed with the results of their AI deployments, and 17 percent would not repeat the decision to implement it. Even so, 77 percent of firms plan to increase AI spending over the next 18 months.
Contents
Nearly half of Polish companies that have implemented artificial intelligence solutions are disappointed with the results. That's according to the latest, third edition of EY's report on AI adoption among Polish enterprises, published on July 12, 2026. The problem, the report's authors stress, isn't the technology itself but a lack of knowledge, organizational readiness and mature deployment processes.
The report's authors note that despite their disappointment with the results, Polish companies aren't giving up on AI investment. Quite the opposite: 77 percent of surveyed enterprises say they intend to increase AI spending over the next year and a half, and a third of them expect a significant budget increase. That suggests companies see current setbacks as a learning stage rather than a reason to abandon the technology.
Where the Problem Lies
According to the report, the key barrier isn't the availability of AI tools themselves but companies' organizational readiness to use them. Only 9 percent of enterprises have the complete data infrastructure needed to build advanced AI-based solutions. As many as 39 percent of companies cite security and cyberthreat concerns as the main obstacle, and 22 percent admit they lack access to experts who could lead a deployment.
A significant share of companies are still at the early stages of AI maturity. On EY's scale, 67 percent of enterprises rank at the third maturity level or lower out of five, and only 10 percent are genuinely transforming their business model through AI.
Differences Between Industries
The report reveals clear differences between sectors of the economy. In financial services, 72 percent of companies sit at the second or third maturity level, making it one of the more advanced sectors. In energy, 21 percent of companies are advanced AI users while 42 percent are still at an early deployment stage, and this sector reports the highest security concerns, with 46 percent flagging them. In manufacturing, 29 percent of companies reached the third maturity level, while in retail only 23 percent use AI in key business processes, most often through isolated, one-off solutions.
Artificial intelligence is increasingly becoming a test of business maturity, not an experiment - Piotr Ciepiela, head of EY Poland Technology Consulting
The Last Mile of Deployment
Piotr Ciepiela of EY Poland points out that technology itself accounts for only a small part of an AI project's success. In his view, it's the final stage of deployment, actually embedding the tool into teams' daily work, that makes up roughly 90 percent of a project's total effort. Many initiatives fail not because of limitations in the AI models themselves but because of poor change management and a lack of employee buy-in.
Among companies that did benefit from their AI deployments, 53 percent cite cost reductions, 52 percent point to improved service quality, and 49 percent report revenue growth. That shows that where deployment went well, the results are measurable and span multiple areas of a company's operations.
Poland Versus Europe
Data cited alongside the report show that Poland lags behind Western Europe in business AI adoption. In 2025, just 8.4 percent of Polish enterprises with at least 10 employees used AI solutions, compared with an EU average of around 20 percent, and as high as 40 percent in Denmark and Finland. That puts Poland among the EU countries with the lowest AI adoption rates.
For Polish managers and IT departments, the EY report signals that AI deployment success depends more on organizational preparedness, data quality and change management than on the choice of a specific tool or technology vendor. Companies that want to avoid disappointment should invest first in data infrastructure and team skills before committing to costly rollouts of advanced models.
Sources: Bankier.pl (bankier.pl), EY Poland (ey.com)
