What happened
A Gartner survey of 2,002 CIOs and technology executives found that only 13% said their enterprise had realized significant or transformational value from AI. The probability that an individual AI use case succeeds remained approximately one in five.
The problem is not simply that experiments fail. Many organizations lack rules for dealing with failure. Only 31% had guidelines to prevent an unsuccessful pilot from moving into production, and just 17% had criteria for abandoning low-value AI already in production. Meanwhile, 58% faced greater pressure to deliver cost improvements from AI, while 51% expected AI to increase the total cost of ownership of technology solutions.
The survey, conducted from May through July 2026 and published October 19, is broad evidence rather than a report of one company’s breakdown. Its organizational-design finding is especially stark: only 4% of CIOs said they had designed roles for humans and AI agents to work together. Organizations with those plans were 2.4 times more likely to be classified as high-performing AI “vanguards.”
Gartner also found that organizations with pilot stop rules were 2.3 times more likely to be vanguards, while those with criteria for abandoning low-value production initiatives were 2.6 times more likely. A second report of the findings is available from Express Computer.
Why it matters
For a small or midsized business, the lesson is less about copying enterprise governance and more about protecting scarce management attention. An AI project can look inexpensive at launch while creating recurring costs for software, review, exception handling, security, and process rework.
Before funding an AI workflow, define:
- The business outcome and current baseline
- Total operating cost, including human review
- The threshold for moving from pilot to production
- The threshold for retiring the system
- Which human work remains, changes, or is added
- One accountable business owner
Buying an agent is not the same as designing a workable role around it. If the workflow, decision rights, and economics are unclear, scaling the tool mainly scales the ambiguity.