Forcepoint

Data Classification
powered by Getvisibility

Combining generative AI and Large Language

Models to deliver highly accurate data classification
for Forcepoint Enterprise DLP and Forcepoint ONE

Brochure

forcepoint.com
Forcepoint Data Classification powered by Getvisibility forcepoint.com

Fundamental Challenge to Data Security

When it comes to an organization preventing the loss of
data, the common phrase, “you can’t protect what you
can’t see,” speaks to the very foundation of data security.
Organizations that lack an effective method for
identifying or “seeing” all the types of data that they
have are ultimately unable to prevent the loss of that
data. However, typical classification methods are highly
manual, largely depending solely on individuals to make
important decisions on the value of data. Inaccurate data
classification generates large volumes of false positives
and false negatives in Data Loss Prevention (DLP)
policy enforcement. This results in wasted time and
resources. It also impacts an organization’s ability to keep
sensitive data from being exfiltrated and leaves it
open to security threats from both outside (malware,
ransomware) and within the organization (malicious
insiders, misuse, error, mistaken classification).

Figure 1: AI-powered 50-dimensional model

Impact of Inaccurate Data Classification Leveraging AI to Gain Sight into

All of Your Data
Forcepoint is delivering an innovative type of data

$4.5
Up to classification technique that leverages the power of Artificial

Million
$20 Million
Intelligence (AI) and Large Language Models (LLMs) to
improve data classification accuracy, automating the
discovery and classification process.

Moving beyond traditional machine learning, Forcepoint

or 4% annual Data Classification employs TensorFlow to create Generative
Average cost of
a data breach* global turnover AI models, constructing a predictive and self-learning
non-compliance fines classification engine. Forcepoint Data Classification works
(GDPR) with natural language to suggest classifications. Using a
50-dimensional model, customer meta-data is ingested
into the model and, through continuous learning, delivers
$600k
per year
high-accuracy suggestions.

To start, thousands of newly created files and emails

are classified for customers. Each file is converted into
a mathematical vector with a classification label that is
False positives drain anonymous and not able to be reverse engineered so
time and manpower that the information can be securely stored. In this way,
(avg. annual cost of Forcepoint Data Classification leverages classification
InfoSec team)** from customers across various industries to build and
* Source: IBM Security: Cost of a Data Breach Report 2023 continuously improve the 50-dimensional model. As an
** The Quest for a Single Set of Unified DLP Policies (IDC)
individual organization deploys and runs Forcepoint Data
Classification, they are also able to maintain their own
50-dimensional model that can learn based on their
unique organizational classification requirements.

2
Forcepoint Data Classification powered by Getvisibility forcepoint.com

Forcepoint Data Classification achieves unparalleled accuracy by combining synthetic and real-world data:
› Leading LLMs generate millions of diverse sample files, ensuring a comprehensive training dataset that spans a vast range of
content types, industries, and scenarios.
› This synthetic data seamlessly complements anonymous customer data, training the 50-dimensional model to learn and
classify with remarkable accuracy.

Forcepoint Data Customer

Classification (AWS) Network

Forcepoint Classification Customer Documents

Optional
Anonymized
File Signatures
Signature Upload
Anonymized Binary File Signature
Encrypted Binary Data Only API

File Signatures

Classification Model + Generative AI

Model Training
Classification
Model Repository

File Classification

Classification
Improved
Results Database
Models

Reports

1. Server Classification Review UI

2. Endpoint User Classification

During file creation, Forcepoint Data Classification The outcome is Forcepoint Data Classification delivers
enables you to add visual labels and metadata, giving “Classification Everywhere,” providing on-the-fly
you the option of selecting your own classification or classification of newly created data anywhere your
using the AI-generated recommendation. people work. Through continual learning and
improvement, it delivers the most accurate classification
Each recommendation is delivered with a confidence tailored to your business needs, all while requiring
level based on the data within the file compared with the minimal effort to deploy and ensuring a superior Total
AI model to help make highly accurate classification Cost of Ownership (TCO).
decisions. The decision then enables you to add visual
labels and metadata to facilitate policy enforcement and
regulation compliance. As the tool is used, it continues › AI-powered for industry-leading precision
to learn and improve. The customization configuration and efficiency
wizard options include:
› Seamlessly deploy with no user training
› Compliance › Increased productivity and reduced
› Classification false positives
› Enforcement rules
› Visual labeling
› Email sharing rules
3
› Exceptions
Forcepoint Data Classification powered by Getvisibility forcepoint.com

Seamless Industry-Leading Precision and Efficiency

Forcepoint Data Classification is an industry-leading solution that seamlessly sits in the Microsoft application ribbon,
providing easy access to the current document classification status. Upon selection, users receive prompts to choose
classifications with AI-powered suggestions. After saving, the chosen classification level is automatically applied and
embedded in the document’s metadata.

Accuracy User Activity Reporting

Forcepoint Data Classification uses advanced AI to
The administrator dashboard lets you track user
suggest classifications to users, dramatically increasing
classification activity, pinpointing areas for policy
accuracy and reducing human error.
updates, employee training or potential insider threat
investigation.

Automation
Compatibility
Automate the classification and protection of your data
in real-time as your teams work and collaborate on apps Seamlessly integrates with a wide array of applications,
anywhere. including the Microsoft Office suite and more, and pairs
with Forcepoint Enterprise DLP and Forcepoint ONE,
enabling immediate AI-powered data classification upon
deployment.

Visual Labeling
Customizable visual labeling allows you to add custom
headers, footers or extra text fields, based on your
organization’s policies.
4
Forcepoint Data Classification powered by Getvisibility forcepoint.com

Streamline Compliance
Forcepoint Data Classification makes it easy for organizations
to meet compliance standards. The platform simplifies the
compliance process with a user-friendly customization wizard,
ensuring adherence to regulatory requirements. Built-in selections
for major regulations further simplify compliance audits and
reporting. Continuous data monitoring ensures ongoing adherence
with evolving regulations so your organization can focus on
innovation and business outcomes.

› Out-of-the-box key regulations

simplifying compliance
› Continuous data monitoring
ensures ongoing adherence to
evolving regulations
› Granular policy controls allow
you to tailor-fit to specific
compliance needs
› Streamlines compliance with the
broadest coverage of data types
in the industry

For more information schedule a demo today.

5
Forcepoint Data Classification powered by Getvisibility forcepoint.com

Appendix A - Key Features:

FEATURE OVERVIEW

Windows and UNIX File Servers, Windows Unix and MacOS

Data Classification computers, Android and iOS mobile devices, NAS, Microsoft Online,
DropBox, Box and GDrive (uses AIbased classifier).

AI named entity recognition models which identify PII based on the

Personal Identifiable Information (PII) Identification text content, offering the best PII identification accuracy available on
the market.

AI for email content and attachments classification. Allows keeping

Email Control
track of communication that involves sensitive subjects.

Monitoring and recording of events related to file activities and

Monitoring
changes to active directory AI anomaly detection.

Widgets provide a high-level overview with the ability to drill down

and conduct detailed forensic investigations. All the findings can be
Reporting
exported in various formats. Generates management executive reports
with high-level overview of data risk posture.

Supports more than 50 file types (including Microsoft Office file types)
including the following standard files types: pdf, doc, dot, xls, xlt, ppt,
pps, docx, docm, dotm, dotx, xlsx, xlsm, xlst, xlsb, pptx, potm, potx,
File Type Support
ppsm, pptm, ppsx, vsdm, vsdx, vstx, vss, vssm, vst, vstm, vssx, dwg, dxf,
jpg, jpeg, png, mp4, jpe, bmp, wav, mov, avi, m4p, m4v, gif, tiff, tif, zip,
mpp.

Staff can classify and tag documents using the same rules as when
they are online with the same warnings, blocking of risky activities and
help assistance to explain the reasoning for the restrictions. While
Internet Dependency
the ML classification suggestions are only available when online, the
pattern-based suggestions are continuously available offline as well
as online.

Out-of-the-box integration with Forcepoint Enterprise Data Loss

Integration Capabilities Prevention (DLP) and Forcepoint ONE Integrated DLP. It also is able to
be used with other leading DLP solutions in the industry.

Out-of-the-box integration is available with the main RMS providers.

Rights Management Partner Integration
This includes Microsoft Azure RMS, Seclore, SealPath and Ionic.

Visual marking and metadata is fully customizable supporting different

Visual Marking Customization
attributes and variables.

Visual markings are flexible in terms of position. Values can be inserted

Existing Header/Footer Management in existing headers/footers or can replace old headers/footers if
needed. Watermarks are also supported.

Reporting does not require additional resources to be enabled on

the client machine that could lead to reduction in performance.
Comprehensive Reporting Logging is stored on a centralized server. Forcepoint Classification will
provide reports on user activity, data at risk, risk scoring and risk by
department.

Email Subject Marking Fully supports subject marking on emails.

Fully supports content checking on embedded as well as attachment

Content Checking on Embedded and Attachment Files
files including ZIP files.

6
 About Forcepoint
forcepoint.com/contact Forcepoint simplifies security for global businesses and
governments. Forcepoint’s all-in-one, truly cloud-native platform
makes it easy to adopt Zero Trust and prevent the theft or loss of
sensitive data and intellectual property no matter where people are
working. Based in Austin, Texas, Forcepoint creates safe, trusted
environments for customers and their employees in more than 150
countries. Engage with Forcepoint on www.forcepoint.com, Twitter
and LinkedIn.

©2024 Forcepoint. Forcepoint and the FORCEPOINT logo are trademarks of Forcepoint.
All other trademarks used in this document are the property of their respective owners.
[FP-Forcepoint Data Classification-Brochure] 26Mar2024