Security Risks in Tool-Enabled AI Agents: A Systematic Analysis of Privileged Execution Environments

Hardik Goel

Abstract

Tool-enabled AI agents are increasingly deployed in cloud-hosted environments and offered as services, where they perform side-effecting operations through privileged tools within execution environments. While such agents enable powerful automation, the security implications of hosting autonomous agents in privileged execution environments are not yet fully explored. This paper presents a structured analysis of security risks associated with cloud-hosted AI agents. We introduce a taxonomy of ris

Categories

Cite This Resource

@article{llmsec202600170,
  title = {Security Risks in Tool-Enabled AI Agents: A Systematic Analysis of Privileged Execution Environments},
  author = {Hardik Goel},
  year = {2026},
  url = {https://arxiv.org/abs/2605.09721},
}