Netskope Introduces Netskope Skylight Agent Action Control to Stop High-Risk AI Actions

Netskope said it renamed its AI security portfolio to bring its products under the Netskope Skylight name and added Netskope Skylight Agent Action Control to classify AI agent actions and stop selected high-risk actions before execution. The change targets how security teams govern what agents are permitted to do.

Netskope said Agent Action Control arrived as part of its move from Netskope One AI Security to a Skylight family of products. Netskope also cited reported figures that it said relate to organizations’ ability to stop risky agent actions and to past AI agent security incidents.

Agent Action Control classifies every action an AI agent attempts into one of nine intent-based categories prior to execution, including access control changes, configuration changes, credential and secret manipulation, data destruction, infrastructure provisioning, potential data exfiltration, potential external communication, remote code execution, and source code change. The capability applies block, allow, or alert based on low, medium, high, or critical risk categories, and it logs actions for policy alerts. Netskope said it runs on network traffic the Netskope platform already inspects, without adding a separate tool or console.

Netskope said the capability includes policy profiles that attach to specific agents so different agent types can operate under different rules, including options to send default or custom notifications to end-users when actions are blocked. Netskope said Agent Action Control is planned to be available at the end of the current quarter.

“Agent Action Control puts a decision in front of every action an agent takes, so a team can say yes to agentic AI without saying yes to the one action that could cost them a production system,” John Martin, Chief Product Officer, Netskope, said. “These hardened, policy-based, deterministic controls are the backstop that prevents AI agents from causing an enterprise incident,” Dr. Grace Trinidad, Research Director for AI Security and Trust at IDC, said.