This paper presents key findings from a global primary research survey conducted by independent firm Dimensional Research. Through this research, SailPoint aimed to examine the current use, adoption, and governance of AI agents, with a particular focus on the distinct risks their identities present compared to those of human and machine identities. It explores issues such as unintended actions, gaps in governance, and the underlying causes of AI agent risk, as well as the extent to which organizations are leveraging identity security tools to provision and manage these identities.