Home / Technology / AI Agents Get New, Multi-Layered Security Shield
AI Agents Get New, Multi-Layered Security Shield
21 Oct
Summary
- Arc AI agent now available, enhancing productivity with security.
- Nvidia's RTX Spark devices power the new AI agent locally.
- Multi-level security infrastructure guards against AI breaches.
ServiceNow and Nvidia have introduced Arc, a personal AI agent designed for daily work tasks and now generally available. Originally announced in May, Arc is built to run efficiently on Nvidia's new RTX Spark devices, supporting local execution on employee desktops.
The agent enhances productivity by integrating with tools like email and Microsoft Teams. It builds memory and skills over time for improved performance and can proactively surface work items by scanning inboxes. Arc prioritizes data privacy by keeping information local and only using frontier models when necessary.
A key differentiator for Arc is its multi-level security framework. This infrastructure moves enforcement beyond software to runtime and hardware layers. ServiceNow's Super Intelligence Control Tower enforces company policies, while Nvidia OpenShell restricts agent behavior, ensuring granular control over accessed systems and data.
AI Passport is an additional security layer that bridges ServiceNow and Nvidia's systems. It manages an enterprise's rules for AI agent access to services, allowing Nvidia's features to intervene if an agent deviates from its permissions. This system continuously monitors agent actions to prevent breaches and unauthorized operations.
