Companies continue to be bullish on AI, with corporations like Meta, Microsoft and OpenAI spearheading the race– but behind the curtains, organizations are looking for ways to balance their IT budgets.
According to a recent UBS survey that polled IT leaders in 125 organizations, a majority (61%) have AI tools and technologies implemented in their businesses across one or more functions, a number that is up 9% since the first half of 2024.
While a 100% of the respondents said that they are on the journey to discover potential AI use cases, ROI remains a point of friction. For many of these companies, the value of AI is ambiguous, and unrealized.
In the report, UBS analysts wrote that only 11% organizations were able to move their AI projects into production within a two-year mark. The rest are struggling, with a majority looking at 2026, or at least 2H ’25, as the timeline when they can move them to production.
Meanwhile, one thing most companies seem to agree on is keeping their IT budgets in check in 2025. Beginning 2024, IT leaders estimated a rise of 5.6% in IT budgets, but this year, that has fallen to slightly under 4.5%.
Although 6 out of 10 CFOs highlight GenAI to be the single most important innovation of our time, PYMNTS Intelligence reports that only half of the firms that use GenAI for five business applications or more say that they have a positive ROI. Companies using GenAI for fewer than 5 applications that report a positive ROI make up a marginal section (6.3%).
Back to the UBS survey, results show that NVIDIA GPUs have gained significant purchase among the responders, going from 50% in May 2024 to 66% in January, 2025. In AI training and inferencing, the GPUs remain the compute platform of choice. Only 17% respondents said they are using AWS’ Tranium chips. Meanwhile, OpenAI continues to hold the leading position in models with GPT 4.0 and GPT 3.5.
Following the GenAI wave, a majority of companies had increased their IT spend in hopes of making big wins. But two years later, half the companies are yet to see ROI, while a small number reports seeing modest ROI.
The bottleneck has left many investors skeptical about seeing any major boosts early in 2025. For now, enterprises seem to be focused on following a handful of unwrit budget planning rules which begin with having a whole budget for IT as opposed to a separate AI budget, and continued allocation of funds for AI projects without making commitments that can draw scrutiny.
“GenAI is acting as a catalyst for the expansion of AI in the enterprise. This creates a window of opportunity for AI leaders, but also a test on whether they will be able to capitalize on this moment and deliver value at scale,” said Gartner director analyst, Leinar Ramos.
