maxkabakov - Fotolia
Content security in today's digital age is a moving target, as the nature of cybersecurity threats and digitized responses for building trust within a digital workplace are continually changing.
But AI can aid in content security at all endpoints.
Content security begins by securing access to a shared repository, and then to particular folders and even to individual files stored within it. An enterprise content management system typically provides security features and functions for maintaining access controls, where only predefined people and groups have permission to access certain sets of files.
Content-based security extends the protection perimeter to particular tasks that occur on the network -- such as restricting who can view, edit, email, download or print a piece of content, or placing a time limit on the access rights and permissions. Team collaboration promotes ad hoc sharing of messages, documents, images and other kinds of content within a predefined workgroup, and it restricts access by outsiders.
These approaches to content security are deterministic, protecting content against predetermined threats. This is akin to locking valuable items inside physical filing cabinets that only those with the right keys can access.
Today's digital workplace demands a greater degree of intelligence and intuition due to the number of endpoints that now exist. Organizations should use cloud services to power content security and embed it within digital experience, whether the content itself is at rest or in motion. AI enhances content security by managing content in context, as well as detecting anomalous patterns to the flow of content across the enterprise.
For example, Box, a premier cloud content management provider, is adding Box Shield to its set of enterprise-scale content security services. Announced in August and slated for release before the end of 2019, Box Shield uses machine learning to detect discrepancies in content usage patterns and enforces classification-based security policies against accidental data leakage.
Box Shield monitors the flow of content among authenticated users within the Box platform, identifies anomalous activities and alerts enterprise security staffers about potential problems, providing them with contextual explanations about suspicious activities. Rather than enforce predefined and inflexible rules, the machine learning algorithms are context-specific -- first setting a user- or group-specific baseline, then flagging events that occur outside the baseline.
Of course, any application of AI must target particular security concerns. Box Shield monitors the activities it can detect and, thanks to machine learning, continually learns about new ones over time.
It's important to note that Box Shield and other approaches to AI-enhanced content security are just one part of the puzzle for building trust into digital experiences. Organizations need an overall enterprise security framework that includes authentication controls to ensure identity, connection controls to protect networked data flows and device controls to lock individual endpoints.
But equally important is an enterprise information architecture where machine learning algorithms use well-defined categories. After all, semantic modeling, enabled through a cycle of metadata enrichment, improves the results of machine learning. The best way to protect content within a digital workplace is to correctly categorize core information assets to begin with.
Dig Deeper on Enterprise content management (ECM) technology
Related Q&A from Geoffrey Bock
Content services platforms are the evolution of enterprise content management systems. Here's what you should know about each technology and how they... Continue Reading
While business process management and workflow work together, they have two very different functions. Learn how they can fit into your business model. Continue Reading
Integrating e-commerce with a CMS can be a challenge, but many vendors are building compatibility into the cloud. Here's how and why those processes ... Continue Reading