Only 28% of AI infrastructure projects fully pay off, survey finds

A survey by Gartner indicates that only 28% of AI infrastructure projects fully deliver a return on investment (ROI), with one in five failing outright. The research, based on responses from 782 IT infrastructure and operations (I&O) managers, found that 57% have experienced at least one AI project failure. Melanie Freeze, a research director at Gartner, attributes these failures to unrealistic expectations, such as assuming AI would immediately automate complex tasks or cut costs. Projects often stall when results are not quickly apparent. Failures are most common in auto-remediation, self-healing infrastructure, and agent-led workflow management. Persistent skill gaps and poor or limited data quality are also significant barriers, cited by 38% of those who faced setbacks. More mature AI applications, like GenAI for IT service management (ITSM) and cloud operations, show higher success rates (53%). The article also highlights challenges in securing funding for AI in tech infrastructure, with many initiatives still funded by individual business units. Gartner advises CEOs and CFOs to take a more active role in setting funding criteria and approving major investments as spending rises. This comes as companies struggle to justify AI spending, with a separate survey showing over 80% detect no discernible impact on employment or productivity despite widespread AI use. Furthermore, tech leaders face increasing pressure from boards to demonstrate AI ROI, with many anticipating budget cuts if targets are not met.

(Source：Theregister)