What’s New in Proactive DLP™ v2.22.0?

Enhanced AI-powered Document Identification and Malicious Insider Threat Detection

Jan 8, 2025 by Stella Nguyen, Senior Product Marketing Manager

Share this Post

Proactive DLP™ is one of several core technologies ( including Deep CDR™ and MetaScan™ Multiscannning ) powering the OPSWAT MetaDefender Platform to prevent sensitive and confidential information in files from leaving or entering organizations’ systems by content-checking files before they are transferred. This aids compliance with regulatory requirements like HIPAA, PCI-DSS, and GDPR. Proactive DLP also uses AI and machine learning to detect adult content in images and offensive language in text.

Better Document Identification with Enhanced AI

Building on its already robust capability, Proactive DLP has further refined its AI-Powered Document Identification feature, delivering even more accurate and efficient document classification. These enhancements enable organizations to better recognize and protect sensitive information across a wider range of document types.

The latest enhancements to AI-Powered Document Identification include:

Enhanced Machine Learning Models

Advanced machine learning algorithms improve detection accuracy, allowing for more precise identification of sensitive documents and reduced false positives.

Expanded Document Type Support

Proactive DLP now supports an even broader array of document types, including:

Financial documents
OCR-processed documents
Specialized file formats

This feature can be activated in MetaDefender Core:

User interface showing options to detect and categorize documents, including financial, energy, and government files, with blocking behavior for sensitive personal documents

Prevent Data Leaks from Malicious Insiders

When it comes to security threats and fraud risks, we often focus on hackers and external attacks attempting to breach companies' networks to steal information and data. However, insider threats are another common source of data leaks.

Threats from employees and other trusted insiders are on the rise. In 2023, the average annual cost of insider security breaches surged to $16.2 million per organization, while the average time to contain such incidents extended to 86 days. More than ever, it’s important for businesses to adopt robust security protocols to safeguard sensitive information and data.

Insider Threats vs External Attacks

Over the past five years, the perception of the challenges in detecting and preventing insider attacks, as opposed to external cyberthreats, has undergone a significant shift. Concern about malicious insiders has increased significantly, rising from 60% in 2019 to 74% in 2024. This rise indicates a greater awareness or experience of intentional insider attacks.

Graph displaying insider threat types—malicious insiders (74%), inadvertent insiders (63%), and negligent insiders (58%)

Insider threats are malicious acts carried out by authorized employees within an organization. These threats represent a major cybersecurity challenge for both private and public organizations, as insider attacks can cause extensive damage to organizational assets, often more than external attacks. Most existing approaches in the field of insider threat detection focus on general attack scenarios. However, insider attacks can be executed in various ways, with the most dangerous being data leakage attacks, which can be carried out by a malicious insider before they leave the organization. This highlights the need for advanced, nuanced methods to identify these emerging risks.

Types of Data at Risk

According to the 2024 Insider Threat Report, financial data, customer data, and employee data are highly vulnerable to malicious insiders. Financial data is particularly at risk due to its direct monetization potential. Customer data follows closely, raising concerns over the loss of PII (personally identifiable information). Employee data is also a significant concern, highlighting risks from mishandling sensitive personnel information. The report emphasizes that all company-sensitive data is susceptible, highlighting the need for comprehensive data security measures.

Bar chart highlighting the most vulnerable data types to insider threats, led by financial and customer data

Detect Malicious Insider Threat Attack Techniques

Malicious insiders often use clever techniques to bypass security measures.

Extremely small font sizes can be used to trick human eyes.

Example report summarizing achievements, new initiatives, team highlights, and upcoming challenges

Invisible text background tactics can conceal sensitive information.

Two versions of a report, one with sensitive text concealed in white font, illustrating the risks of hidden data

Sample of invisible texts inside a sensitive document

User interface for enabling detection of hidden or small-sized text in documents to prevent data leakage

Sample of actual text disguised as invisible text

With Proactive DLP, organizations can proactively detect and prevent these techniques.

Enable this feature in MetaDefender Core:

System notification of detected sensitive data, including invisible text, in a scanned document

Proactive DLP detects sensitive information.

Report identifying sensitive data, including certainty levels and the nature of the hidden information

Administrators can view the details of the detected sensitive information.

This new feature ensures that no malicious insider techniques escape scrutiny, empowering organizations to maintain tighter control over sensitive data and prevent regulatory violations.

Protect Sensitive Data with Proactive DLP

The threat of malicious insiders is a reality that no organization can afford to ignore. By leveraging advanced technologies like Proactive DLP, companies can gain a competitive edge in protecting sensitive information and preventing regulatory violations. As the use of AI-powered document classification and malicious insider threat detection continues to evolve, it's essential for organizations to stay informed and adapt their security strategies accordingly.