87 Percent of Polish Developers Use AI Daily, Universities Can't Keep Up

Polish software developers use artificial intelligence far more often than their counterparts in other countries, and at the same time are broadly critical of universities for failing to prepare students to work with the new tools. That is the conclusion of the DevHorizon Poland 2026 report, prepared by the agency Strand Partners on behalf of Amazon Web Services.

The DevHorizon Poland 2026 study is one of the first large-scale attempts to measure how deeply generative AI has changed the daily work of developers in Poland. Strand Partners surveyed 500 active developers, asking about both their everyday habits and the gaps they see in the education they received at university.

Poland ahead of the global average

The figure of 87 percent of developers using AI daily stands well apart from global data. The Stack Overflow Developer Survey 2025 found that 51 percent of developers worldwide use AI tools every day. The report's authors note that Polish engineers are adopting new tools faster than the international average would suggest.

The reported time savings amount to an average of 7.3 hours a week, and 85 percent of respondents say AI has genuinely boosted their personal productivity. That is close to an entire workday reclaimed every week, which developers are redirecting toward tasks that require deeper thinking rather than routine coding.

Artificial intelligence is changing the developer profession, but the tools alone are not enough - Andrzej Horawa, country manager, AWS Poland

The end of grinding out code

The survey shows the developer's role shifting away from mechanically writing code and toward supervising and verifying what AI models generate. Asked about the key skills they will need in five years, respondents pointed to critical thinking (57 percent), system architecture design (55 percent), and managing AI systems (53 percent). Only 11 percent of respondents still named traditional coding skills as key.

This shift raises questions about accountability for code that a developer did not technically write, but rather generated and approved with the help of an AI assistant. The industry is increasingly noting that the ability to read and understand someone else's code, including machine-generated code, is becoming more important than the ability to write it from scratch.

The industry needs to ask whether the people managing these systems understand the code well enough to take full responsibility for it - Jarosław Dąbrowski, country manager, JetBrains Poland

Universities falling behind

The sharpest criticism in the survey is directed at higher education. 79 percent of developers believe computer science curricula need urgent, faster reform, and 78 percent say outright that schools and universities are completely failing to keep up with the market's transformation. As many as 68 percent of respondents would like AI skills training to begin already in secondary school rather than waiting until university.

The problem is not limited to computer science programs. The report's authors point out that artificial intelligence is no longer the domain of developers alone and is becoming a working tool in energy, medicine, finance, media, and public administration. That means the skills gap respondents are flagging extends to a wider group of graduates than just IT majors.

What it means for the job market

Researchers partly link the decline in job openings for junior developers, observed in Poland for several years now, to AI tools taking over simple, repetitive tasks. At the same time, the report's authors caution that AI is not replacing the profession as a whole, only cutting out the mechanical portions of the work and pushing developers toward higher-level engineering tasks.

For IT companies, this means rebuilding how their teams work, not simply buying licenses for AI tools. The survey found that real productivity gains only appear once an organization restructures its entire software development process around the new tools, rather than when individual developers use them on their own.

For students and universities, the report's conclusions are unambiguous: job market expectations have outpaced rigid computer science curricula, many of which were designed long before generative AI became widespread. Pressure to update curricula more quickly is likely to keep growing as successive graduating classes run into the realities of the market.