Netskope Launches AI Agent Action Control to Mitigate Risks

• •Netskope launched Skylight Agent Action Control to prevent risky AI agent actions.
• •The system categorizes actions into nine types and allows for risk-based policy enforcement.
• •91% of organizations cannot stop risky actions before execution, highlighting the need for this tool.

Netskope has introduced the Skylight Agent Action Control, which allows security teams to classify and block high-risk actions taken by AI agents before execution. This new capability is part of the rebranded Netskope Skylight AI Security suite, which includes various products aimed at enhancing AI security. The system categorizes agent actions into nine intent-based types and supports risk-based policy profiles tailored to specific agent types. This launch addresses a significant gap, as 91% of organizations reportedly cannot prevent risky agent actions before they occur. The new controls operate on existing Netskope-inspected traffic without needing additional consoles or agents. The availability of this feature is expected by the end of the current quarter. Following the announcement, Netskope's stock experienced a slight decline of 0.68%.