According to a new report from Gartner, the effectiveness of deploying artificial intelligence (AI) within IT infrastructure is falling well short of market expectations. Among the 782 IT infrastructure and operations (I&O) managers surveyed, only 28% reported that their AI projects had achieved a full return on investment (ROI).
The survey indicates that one-fifth of I&O AI projects end in failure. Furthermore, 57% of respondents admitted to experiencing at least one failure when applying AI to their own operations.
Gartner Research Director Melanie Freeze notes that many AI projects falter because businesses harbor unrealistic expectations. "Organizations often assume that AI will immediately automate complex tasks, slash costs, or solve long-standing operational headaches," she said. "When reality fails to meet these expectations, confidence wanes and projects stall."
The Root Causes of Project Failure
Gartner’s research found that AI failures in the I&O space are concentrated in areas like automated remediation, self-healing infrastructure, and agent-driven workflow management. Freeze added that the 20% failure rate is largely driven by projects that are overly ambitious or poorly defined. If AI cannot be seamlessly integrated into an organization's core operational processes, it is destined to fail to produce tangible benefits.
The technical talent gap remains another major hurdle. Among respondents who encountered project setbacks, 38% blamed the persistent skills shortage, while an equal percentage pointed to poor data quality or limited data availability as direct causes of failure.
Despite the lackluster overall returns, some progress has been made in more mature technical areas. Data shows that 53% of I&O leaders have achieved success by applying generative AI to IT Service Management (ITSM) and cloud operations.
Currently, funding for AI projects remains a challenge. Freeze observed that many AI initiatives are funded independently by individual business units. As AI infrastructure spending continues to climb, CEOs and CFOs must take a more active role in approving major investments and setting funding standards.
These findings highlight the widespread struggle companies face in justifying AI expenditures. A survey conducted in February among nearly 6,000 executives across the U.S., U.K., Germany, and Australia found that while 69% of companies are using some form of AI, more than 80% of respondents have yet to see a significant impact on employment or productivity. With boards increasingly demanding proof of ROI, many CIOs face the risk of budget cuts or freezes if their projects fail to meet expectations by the first half of this year.
