Artificial Intelligence
Written by: Tathagata Sen
Updated 12:58 AM EST, November 17, 2026

Only 18% of Chief AI Officers (CAIOs) have full control over AI investment decisions, while just 29% say their organizations quickly adjust resources when performance data identifies opportunities to improve results, according to a new IBM Institute for Business Value study published November 16.
The study, conducted with Oxford Economics, surveyed more than 1,600 CAIOs from April to June 2026. Respondents were asked about AI strategy and priorities, investment management across the AI roadmap and the operationalization of AI.
The findings point to a gap between the responsibility placed on AI leaders and their authority to act on what they learn about AI performance. Although these executives may have visibility into how AI initiatives are performing across the enterprise, investment and resource decisions often remain with other parts of the organization.
IBM found that many organizations have AI governance frameworks in place. While 73% say they have defined where human oversight is non-negotiable, only 61% report that AI decision rights are clearly defined and consistently applied.
The investment gap is equally pronounced. Forty percent of respondents say decisions are discussed when performance insights reveal opportunities, but action is delayed. Only 29% say their organizations adjust decisions and resources quickly in response to those insights.
IBM identified a group of high-impact AI leaders whose organizations report stronger results. These leaders deliver AI return on investment (ROI) at more than double the rate of other CAIOs, according to the study.
These organizations also appear more willing to redirect resources when initiatives fail to deliver. Ninety-four percent of high-impact leaders say they use business outcomes to determine which AI initiatives receive additional funding and talent. And 96% say they stop or reduce investment in initiatives that fail to meet expectations, even when they have senior sponsorship. Among other respondents, that figure is 46%.
The report also associates these practices with stronger performance across AI initiatives, including better business outcomes, more pilots moving into production and faster enterprise-wide scaling.
For chief data officers (CDOs), the findings highlight the importance of connecting data foundations, AI governance and investment decisions. AI leaders need reliable performance data to assess whether initiatives are delivering value, identify where models and workflows need adjustment, and make a credible case for reallocating resources.
That responsibility also extends to decision rights. As organizations expand the use of AI agents across business workflows, they need clear rules for which decisions can be automated, where human oversight is required and who can intervene when systems behave unexpectedly. These are increasingly central concerns in enterprise AI governance.