AI NewsGo-to-marketAnnouncement

IBM says 18 percent of Chief AI Officers control the AI budget

IBM surveyed 1,667 senior AI executives in 33 countries. It found 18 percent of them have full control of the AI budget they are judged on.

AI News

Editorial3 min read

LinkedInX
Banner image from IBM's Chief AI Officer study release

Image: IBM

Why it mattersA buyer of AI work is often not the budget holder. A vendor that talks only to the Chief AI Officer may miss the person who signs.

A new IBM study says most Chief AI Officers can see where AI makes money in their company. Few of them control how that money is spent. IBM says 18 percent of the senior AI executives it surveyed have full control of AI investment decisions. The other 82 percent do not.

The study was published by IBM's Institute for Business Value on 17 November. IBM ran it with Oxford Economics. They surveyed 1,667 senior executives who lead AI at their company. Titles included Chief AI Officer, Chief Data and AI Officer, and Head of AI. The survey ran from April to June 2026. It covered 33 countries and 20 industries.

The headline numbers

All figures are IBM's own. 18 percent of the surveyed executives said they have full control of AI investment decisions. 29 percent said their organisation quickly adjusts decisions when new insights appear. 48 percent said AI agents already act across workflows at their company. 75 percent said governance, risk and compliance delays slow AI work down.

IBM also splits the sample into a group it calls high-impact leaders. 63 percent of projects from that group deliver measurable outcomes. The figure across the full sample is 43 percent. IBM does not define the cut-off between the two groups in the release.

What the lead author said

Joanne Wright is Senior Vice President for Transformation and Operations at IBM. She said in the release: "Few leaders have a broader view of how AI is creating value across the enterprise. Yet the outcomes they are accountable for often depend on decisions made by others."

What this changes for a buyer or seller

A software team selling AI work to a large company often asks for the Chief AI Officer. That title now reads as the right person to pitch. The IBM survey says that in most companies the title does not come with the budget. The person who signs for the work often sits in a different function. A seller who maps the real approver before the pitch will spend less time on introductions. Those introductions do not lead to a contract. A buyer inside the Chief AI Officer's team should bring that person along early. They will need them when the pilot moves to a real line item.

The limits of the number

One number from a survey is a claim. IBM's sample is self-selected. Only companies with an executive dedicated to AI could answer. The 18 percent figure is the share of those executives who report full control. The other 82 percent sit somewhere between partial control, shared control, and no control at all. IBM does not break those apart in the release. A reader who wants the full breakdown will need the longer report that IBM says is out today.

The agentic-AI number is also worth reading carefully. 48 percent of respondents said agents initiate and execute actions across workflows. That is a self-report. It does not say the agents are good, or that they are trusted with high-stakes work. 75 percent of the same respondents said governance delays slow AI work. A fair read is that many companies have agents running today. Those same companies are still working out how to keep a hand on them.

Source

IBM Newsroom, "IBM Study: Chief AI Officers See Where AI Creates Value, but Few Control Investment Decisions".

This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.

Share
LinkedInX
Start a project