Death, Taxes and Redaction Blunders

Avoiding eDiscovery redaction failures 

By Rachel Teisch

Nothing in this world is certain except for death, taxes and redaction failures. Recently, The Register discovered a court document related to an age discrimination suit filed against IBM in Texas last year in which the plaintiffs replied to IBM’s partial motion to dismiss the case because some plaintiffs don’t reside in Texas. The plaintiffs failed to correctly redact the documents and disclosed IBM documents that support a denial of its jurisdictional challenges—disclosing information to the public that was previously confidential or unknown in the process.

Redactions are harder than you think

Redaction, in the context of litigation, is the process of obscuring text in a document because that text is either confidential, privileged or otherwise exempt from disclosure. Redactions should look and act like a permanent marker that forever blacks out sensitive passages. In fact, prior to digital redaction methods, a black marker was how it was done. Lawyers printed out paper copies of emails and other electronic documents, used a black marker to “x” out sensitive passages, and scanned the documents for electronic productions.

Today, most lawyers use digital tools to redact. Digital tools offer speed and convenience. But if you don’t use them properly, you can wind up embarrassed—if you’re lucky—or face more serious consequences if you’ve really fouled up.

If the redaction isn’t “burned” into the image so that it cannot be removed, the content underneath can still be viewed. When using Adobe Acrobat PDF files, for example, the most common mistake is to obscure the text by drawing a black box over the text or images you want redacted. However, a few clicks easily remove the black box.

Even Acrobat’s redaction tool isn’t foolproof; redaction annotations can be easily removed and then the content underneath them is revealed, and the underlying text can be viewed by copying the passage and pasting the results into a word processing application. That’s what happened in the IBM filing—the redaction didn’t take.

Redact with confidence

Advanced eDiscovery platforms should allow you to redact with confidence. They also ensure consistent redaction designations across similar documents, save you substantial time through automation, and also help you safeguard sensitive PII to meet data privacy compliance obligations. It is essential that the redacted data is physically deleted and not just masked.

Redaction features in every lawyer’s toolbox should include:

  • Support for both document-by-document (discrete) and bulk redaction;
  • Automated detection of sensitive PII such as date of birth, SSN or SIN, email addresses and states;
  • Custom redaction for virtually any pattern, string or search term through regular expression (RegEx);
  • Tools to redact selectively within documents to specific pages or across all pages;
  • Ability to “undo” redactions by reviewers or case managers with sufficient permissions (prior to finalizing the redactions via productions);
  • The ability to remove the underlying text behind the redactions to prevent exposure when users copy and paste; and,
  • Audit trails for reviewers and case managers to see who redacted exactly what in case there are questions about what was or was not redacted.

What about those frustrating spreadsheet redactions?

Microsoft Excel presents particular complexities when it comes to redaction, and many eDiscovery platforms haven’t solved this to perfection. There could be hundreds or even thousands of pages of columns and rows which need to be lined up to be properly redacted and reviewed. That’s a lot of room for missing essential data that may be concealed in formulas, comments, or hidden rows. These all may be lost once the Excel document is converted to an image.

Fortunately, Excel files can now be redacted in a native-like experience to solve the challenges, provide efficiency gains and ensure more accurate productions.

Redaction tools for Excel include the same pre-defined string types for other types of data (e.g., SIN, date of birth, email addresses, state) and is augmented by custom RegEx for any pattern or search term of interest. Data can be redacted by cells, rows, columns, worksheets or workbooks, either individually or in bulk.  

Advanced eDiscovery systems will also support highlighting to pre-stage redaction for quality control (QC) during review, along with automated QC to check that everything staged for redaction has been properly redacted during the production process.

Don’t forget redaction for data privacy compliance While the ability to redact sensitive data is one thing, finding everything that needs to be redacted is a different, larger problem. When it comes to data privacy compliance, it is important to redact the names of people and organizations while finding the data associated to them—that may also need to be redacted. Now, thanks to the power of automated entity detection, in conjunction with bulk redaction capabilities, you can efficiently and completely search a data set and cut the risk of exposing the personal data of third parties.

Rachel Teisch is Senior Director of Product Marketing at OpenText Discovery. She brings nearly two decades of experience in eDiscovery, and is responsible for product marketing for the OpenText Discovery suite of products. She most recently served as Vice President, Marketing, at Catalyst Repository Systems, which was acquired by OpenText in January 2019 and is now part of the OpenText Discovery portfolio.


  • TEAM OpenText™

    OpenText™ is a Guardian Trusted Partner of EDRM. OpenText™ delivers the competitive advantage to corporate legal departments and law firms with end-to-end eDiscovery software and services that lower cost, risk and inefficiency at all stages of the EDRM workflow. See their partner page here:

    View all posts