A new global study by EPAM Systems is calling time on AI hype, revealing that many businesses claiming to be ahead of the curve in artificial intelligence are struggling to walk the talk.
Based on responses from over 7,000 participants across nine countries and eight industries, the report titled From Hype to Impact: How Enterprises Can Unlock Real Business Value with AI found a major gap between how advanced companies say they are with AI and what they’ve actually delivered.
Nearly half (49%) of respondents rated their organisations as “advanced” in AI, with a bold 5% calling themselves “disruptors.” However, only 26% of these supposed leaders have successfully launched real-world AI use cases.
That gap doesn’t stem from a lack of ambition. Companies plan to boost AI spending by an average of 14% in 2025, with disruptors expecting more than half of their profits next year to come from AI. But the reality is that many are stuck in the experimental phase, unable to scale. Just 30% of tech-forward organisations have managed to implement AI at scale, while others remain bogged down by outdated infrastructure, poor coordination between business and IT, and a lack of governance frameworks.
Legacy tech is holding back progress
EPAM’s report argues that businesses still treat AI as something to bolt onto existing systems rather than as a catalyst for broader transformation. It suggests that real progress will come when companies align people, data, and technology around a clear business purpose.
Among the biggest stumbling blocks are talent and leadership. While 65% of disruptors claim to understand the skills required for AI success, that understanding doesn’t always translate into results. Meanwhile, 31% of executives blame legacy technology for holding back AI adoption, though EPAM says the deeper issue is the disconnect between strategic and technical teams. Even companies with the right tools struggle to modernise effectively when their business goals and engineering roadmaps are out of sync.
A third of businesses say weak security programmes are blocking progress, with cloud security, data quality, and data protection high on the worry list. And while 75% of advanced companies claim to have an AI strategy, only 4% of disruptors have established a proper governance framework, despite acknowledging it typically takes 18 months or more to build one.
The findings echo predictions from Forrester, which sees 2025 as a year of renewed strategy, deeper partnerships between business and IT, and a pivot back to predictive AI, all underpinned by better governance.