AI Use in Companies: Wishful Thinking Meets Reality
In the News
AI Use in Companies: Wishful Thinking Meets Reality
A study by EPAM shows that almost half of companies consider themselves advanced in AI—but only a few have market-ready applications. What's going wrong?
According to the EPAM study, almost every second company rates itself as progressive or even disruptive in their use of AI. Realistically speaking, however, only 26 percent of these companies have launched concrete AI products in the market. While AI is widely celebrated as a productivity booster, scaling often turns out to be its Achilles heel: Only 30 percent of "progressive" companies achieve widespread use of AI. Clearly, there's a huge gap between wishful thinking and practice.
Productivity first, customer loyalty second?
The study, for which EPAM surveyed over 7,300 decision-makers and developers in nine countries, reveals that companies primarily seek quick wins. Goals such as "greater productivity" and "operational efficiency" rank far above longer-term goals such as revenue growth or improved customer experience.
In particular, the neglect of customer loyalty could take its toll, as sustainable competitive advantages are increasingly defined by excellent customer experiences.
On average, companies plan to increase their spending on AI by 14 percent in 2025, highlighting the growing awareness of strategic AI investments.
Humans as an underestimated AI factor
An ironic finding: The more technological AI becomes, the more important humans become. While 43 percent of companies plan to hire AI specialists – especially machine learning engineers and AI researchers – at the same time, 42 percent see a massive need for upskilling among existing staff.
The AI job engine is thus in strong motion, as almost every second company wants to create new AI-related positions. However, many companies apparently underestimate that AI success depends primarily on qualified and integrated personnel – and not solely on the technology.
Security and governance lag behind
The fact that only four percent of companies considered disruptive have established comprehensive governance structures for AI highlights another serious weakness. On average, companies need a full 18 months to establish such structures.
The situation is similarly problematic with IT security: 35 percent of companies complain that inadequate security programs are slowing their AI adoption. This could become a stumbling block for many, especially given increasing regulatory requirements, such as the upcoming EU AI Regulation.
See the original article published in German here.
To read the full 2025 Artificial Intelligence Report, visit: www.epam.com/ai-report-2025.
Learn how EPAM helps drive AI strategy, delivery and enablement here.
