Our Insights

Thought Leadership and Industry Trends

Home 9 Insights 9 AI 9 The Role of Legal Technology in Managing Sensitive Data

The Role of Legal Technology in Managing Sensitive Data

Oct 10, 2024

Finding, redacting, and protecting sensitive data used to be—and in many cases still is—a manual process for hands-on attorney reviewers armed with black markers and a sharp eye for detail. However, surging data volumes, tight deadlines, and ever-changing regulations have required using advanced eDiscovery software and tools for identifying, redacting, and protecting sensitive data.

Planning for sensitive data during eDiscovery and preparing the organization and legal teams before litigation takes place is essential for avoiding any legal ramifications or reputational or economic harm. This due diligence will provide the time to better understand the obligations required and what may be necessary to address them.

Benefits of AI and Machine Learning in Sensitive Data Detection

Some standard methods of finding and protecting sensitive data utilize search terms, clustering, technology-assisted review (TAR), and continuous active learning (CAL). However, the sheer volume and complexity of data render traditional review protocols and workflows inadequate in protecting sensitive information. In response, cutting-edge companies are turning to artificial intelligence (AI) and machine learning (ML) technologies to strengthen their data security strategies.

Artificial intelligence and machine learning provide significant benefits over traditional methods for those charged with protecting sensitive data, including:

  • Resource savings and scalability
  • Superior accuracy
  • Fewer false positives and negatives
  • Seamless behavioral analysis
  • Quicker response times

In the future, look for Generative AI models that can help identify documents potentially containing sensitive information and even explain why they might be considered sensitive.

Tools to Identify and Protect Sensitive Data

Once data is securely within a review platform, legal technologists have a wealth of technology and workflow options to help identify and protect sensitive data during disclosure and production. For example, Relativity Redact, the most comprehensive automated redaction software tool, allows users to reduce review time, lower costs, and increase accuracy by automating redactions of sensitive information based on search terms or regular expression.

Recently, Relativity introduced multiple cloud-based machine-learning tools that we have successfully paired with various CDS Vision dashboards to identify and protect sensitive data. Utilizing indexes, search terms, categories, and automated workflows, these tools are more powerful when used together.

CDS Vision Sensitive Term Widget + Relativity Sentiment Analysis

Our proprietary CDS Vision dashboards allow clients to easily pinpoint and address sensitive data across collections. Relativity Sentiment Analysis, first introduced in 2023, uses Azure-based machine learning to detect and rank the strength of four sentiments: anger, desire, and negative/positive tone. However, occasionally the model produces false positives. Our development team has enhanced Sentiment Analysis with its proprietary CDS Vision Sensitive Terms. When Sentiment Analysis is paired with CDS Vision sensitive terms, data can be sliced in multiple ways, with highlights on both sentiment and sensitive term at the document level.

CDS Vision Personal Information Widget + Relativity PI Detect

Launched in 2023, RelativityOne PI Detect is an AI-powered solution that detects and redacts personal information (PI) with a set of pre-trained detectors. PI Detect uses a combination of rules and machine learning models to identify the context and document structure to reduce false positives, automatically highlights all personal information that has been identified, and generates a document report. When we pair Relativity’s PI Detect with the CDS Vision PI widget, broad categories of personally identifiable information (PII) are easily seen. To use Relativity PI Detect more efficiently, it can be paired with the CDS Vision PI widget to gain insight into a data set beforehand.

CDS Vision Privilege Domain Widget + Relativity aiR for Privilege

The Relativity aiR suite, introduced in 2024, uses generative AI (Microsoft Azure OpenAI GPT-4 Omni). The current aiR products are aiR for Review, aiR for Privilege, and aiR for Case Strategy. CDS pairs aiR for Privilege with Vision, which displays law firm names and domains as well as privilege terms. On its own, aiR for Privilege doesn’t detect privilege for every outside counsel mentioned in a document set. While it’s still a good idea to ask your client for their outside counsel list, CDS Vision Privilege will provide those insights beforehand. To avoid bias, validation of aiR results should happen without looking at the aiR results. CDS Vision Privilege highlights can be used to put the focus on parts of a document that might warrant a longer look.

Boost Your Organization’s Handling of Sensitive Data With CDS

CDS has hundreds of active matters that provide broad insight into end users’ needs and how technologies can be paired to be even more useful. To learn more about elevating your organization’s handling of sensitive data in eDiscovery, contact us at .

About the Author

Deerrun Jea

Deerrun Jea

Deerrun Jea is a CDS Product and Solutions Manager based in New York, focused on Convert and Vision. As a lawyer-technologist, he integrates the practice of law and the impact of technology, including advising both external and internal clients on workflows and strategies related to eDiscovery. He has over 20 years of legal and eDiscovery experience, including multidistrict litigation, FTC/DOJ second requests, and internal investigations.