eDiscovery AI – PII Extract

eDiscovery AI – PII Extract

PII detection and extraction are often some of the more expensive, time consuming and labor-intensive aspects of many document reviews. In addition, with Data Breach reviews becoming more prevalent and wide ranging in the legal technology industry, better and more efficient means of identifying and notifying individuals whose names and information may have been disclosed (often in tight turnaround times) is now more important than ever. 

The inclusion of the new PII Extract solution expands on existing functionality (PII Detect) and incorporates additional associated features for use. 

PII Detect 

The template of optional assorted PII fields has been removed from the Relativity template and is now located in the “Hub” dashboard here. Now, when running a PII Detect or Extract functionality, the application will only specifically detect and extract the types identified through the Hub template dashboard.  

A screenshot of a computer Description automatically generated 

 

PII Extract 

PII extraction is a new feature included with this release.  This identifies named individuals along with various identified descriptors included to the Hub Template and extracts that information to a location within Relativity.  To use:  



  1. Identify the document set you wish to run “PII Extraction” against and select “Sent to eDiscovery AI” from the mass action menu.  

  1. Select the PII as the Job Type.

  1. Select the multi-choice field you wish to use for the PII Detection Results. Then, check the box for "Extract PII."   

  1. Select Next, then select Submit after reviewing the job details.  The documents will be sent to eDiscovery AI and the results extracted.  




Limitations to the use of PII Extract are the same as they are for Relevance and Image review (48k tokens, file size limit).  There is also a cap of 100 names for extraction; if documents contain more than 100 names, there are other options available for us to extract that information which would be less costly. Documents with more than 100 names will be returned as “Technical Issue.”  

eDiscovery AI Extraction Tab/Object 

Upon uploading the updated application, you should see a new tab for “eDiscovery AI Extraction.” This can be seen in Relativity under the “hamburger” menu icon.  


 

Names and descriptors extracted through the PII Extraction icon will accumulate here.  Regarding the viewing fields, once the PII Extraction functionality has been run with the various descriptor fields, those fields can then be added to the view as well.   

 

On a document level, if you go to the eDiscovery AI Layout within the document viewer, you will see the names extracted as part of the view, just below document explanations.  


 

    • Related Articles

    • eDiscovery AI PII Detect

      PII Detection 1. Use eDiscovery AI PII Detect tool to review documents and identify any that contain personally identifiable information. The types of PII considered in each review can be selected and adjusted by the user. a. To conduct a PII review, ...

    • Drafting Effective PII Prompts

      Writing Effective Prompts for PII Detect and Extract Overview This guide covers best practices for creating custom prompts to detect and extract personally identifiable information (PII) from documents using eDiscovery AI. Well-crafted prompts ensure ...