
AI-generated editorial illustration of enterprise AI budgets and investment decisions.
Just 18% of chief AI officers in a new IBM survey fully control AI spending. Most depend on choices made by others. Only 29% say their firms act fast when AI data shows a way to improve results.
IBM shared the survey findings on November 17. The work covers chief AI officers and people in similar roles.
Chief AI officers lack full budget control
These leaders can see how AI works across a firm. Yet other teams may decide which projects get funds. IBM says this can limit their power to act on results.
Fewer than a third say AI is changing their business model. That includes what the firm sells and how it sets prices.
Agents act across more work tasks
Almost half, 48%, say AI agents start and carry out steps across workflows. Three quarters say risk, rule or compliance checks delay AI-driven action.
Many firms have set rules for human oversight. In the survey, 73% say they have defined where it is required. A smaller share, 61%, say rights to make AI decisions are clear and used consistently.
Some firms report stronger results
IBM’s full study singles out a group it calls high-impact leaders. They report that 63% of AI projects give measurable business results. Other leaders put that share at 43%. The first group scales projects in eight months. Other leaders report 13.5 months.
The first group says 49% of pilots reach production. The share for other leaders is 33%. The study links stronger results with habits such as ending failed projects. It also tracks how chief AI officers use data to shift funds.
IBM worked with Oxford Economics to survey 1,667 senior leaders. The work ran from April to June 2026. It covered 33 geographies and 20 industries. These survey links do not prove cause and effect.
Our AI at Work guide covers pilot costs and review time. The Claude for Work guide looks at tasks and tools.
Read also
Popular news

Anthropic opens at $111.50 on Nasdaq, 34.34% above IPO price

GAO tests found safety gaps in four major AI chatbots

Nvidia revenue doubles to $115.8 billion as Rubin ships

Your AI Model Hasn’t Changed. The Company Running It Might Have.

Nvidia RTX Spark mini PCs bring up to 128GB memory to local AI



