Cloud-based software company Box Inc. today launched Box Shield Pro, a new artificial intelligence-powered add-on that expands its existing content security platform with agentic AI designed to automate content classification, detect ransomware faster and streamline security operations.

Box Shield Pro has been built to address the surge in unstructured data driven by AI and automation. The aim is to reduce manual workflows while enforcing security controls without disrupting end users.

The issue the new offering is seeking to address is a growing one – organizations having to balance facilitating and empowering users with keeping sensitive content in only the right hands. Added to the mix is that content is also exploding, with generative AI and AI agents delivering automated productivity beyond anything seen before. According to Box, as a result, security teams are overwhelmed and legacy solutions fall short.

Box Shield Pro allows organizations to secure the growing amount of unstructured content at scale by applying the right controls automatically, detecting threats early and empowering security teams without adding friction for end users.

The new add-on starts with an AI Classification Agent, which uses context-aware AI to automatically label sensitive content based on plain-language prompts defined by organizations.

The agent uses context-aware AI to understand the meaning of documents such as product designs, financial strategies, training materials, meeting transcripts and even creative content like scripts. Doing so allows organizations to move beyond rigid rules and apply security labels with far greater accuracy.

The second part of the offering, Ransomware Activity Detection, extends Box Shield Pro’s protection beyond the cloud into the endpoint, addressing a major blind spot in many content security programs.

The system watches for behavior consistent with ransomware attacks, such as the mass modification or encryption of files originating from a single device. When suspicious activity is detected, administrators receive alert notifications within minutes, including detailed context about the affected user session and impacted files.

The third part of the new offering, AI Threat Analysis Agent, is aimed at reducing security operations fatigue. The agent converts raw threat telemetry into concise, readable summaries embedded directly into Shield alerts.

The AI Threat Analysis Agent delivers summaries that explain what happened, why it matters, who is affected and what the likely impact is all in plain language. With the results, analysts can understand the significance of an incident almost instantly without digging through logs or low-level event data.

Box Shield Pro is available immediately as a paid add-on for enterprise customers using the existing Box Shield content protection platform, while the AI Threat Analysis Agent is scheduled to arrive in early 2026.

Image: SiliconANGLE/Ideogram