Only 13% of CIOs say they've gotten significant value from AI, Gartner finds

Only 13% of technology leaders say their companies have realized significant or transformational value from artificial intelligence (AI), according to Gartner's annual survey of 2,002 CIOs and technology executives, released Monday (Oct. 19) at the Gartner IT Symposium/Xpo in Orlando, Florida. The odds of an AI use case succeeding remain one in five, the research firm said.

The survey, the 2027 CIO and Technology Executive Agenda, was conducted between May and July 2026. It comes as companies rapidly expand their AI projects. Alicia Mullery, a vice president analyst at Gartner, told IT Brew that the average number of AI use cases at companies grew 400% between 2024 and 2025. In its release, Gartner said that although use cases have accelerated over the past two years, the likelihood of each one succeeding remains at that level.

At the same time, pressure for results has increased. According to Gartner, 58% of CIOs report more pressure to deliver cost savings from AI, while 51% expect the technology to raise the total cost of ownership (TCO, which covers purchase, operation and maintenance) of technology solutions. The firm attributes part of the problem to funding and governance models designed for traditional technology investments such as cloud and software as a service (SaaS).

What it means for businesses

The survey highlights a group Gartner calls "vanguards": companies whose CIOs deliver better business outcomes and whose CEOs see high value in AI. What sets them apart, according to the firm, is discipline in choosing and ending projects. Organizations with criteria to stop an AI pilot from moving into production are 2.3 times more likely to be vanguards, but only 31% of CIOs say they have such rules. For abandoning low-value applications already in use, the ratio is 2.6 times, and just 17% have criteria of that kind.

Another issue is how technology is budgeted. Unlike traditional technology costs, which are largely predictable, AI costs are probabilistic and determined at run time, Mullery said. As a result, she said, CIOs will need new approaches to funding, governance and cost accountability. In practice, with tools billed by usage, the bill depends on the volume of tasks performed, not just the number of licenses purchased.

The third issue involves staffing. Companies with explicit plans to redesign roles for a workforce combining people and AI agents (systems that carry out tasks with some degree of autonomy) are 2.4 times more likely to be vanguards, but only 4% of CIOs report having such plans. Contrary to expectations of cuts, 40% of respondents expect to increase IT headcount in 2027, mainly in cybersecurity and in data and analytics. Mullery told IT Brew that Gartner found no difference by industry, company size or IT team size among vanguard organizations.

“Volume is not value.”

— Alicia Mullery, vice president analyst at Gartner, in the survey release

What to watch next

• Exit criteria: how many companies start defining, before launch, the expected value of each AI project and when to abandon it.
• Costs: whether the expectation of 51% of CIOs of a higher total cost of ownership shows up in 2027 budgets.
• AI agents: whether the 4% share of companies with plans to redesign roles for mixed teams of people and agents grows.
• Gartner calendar: the event continues with editions in Yokohama (Nov. 4-6), Barcelona (Nov. 9-12) and Kochi, India (Nov. 16-18).