In the digital age, courts should be more accepting of redactions to protect irrelevant information from production in civil discovery. Redacting irrelevant personal and commercial information poses little risk or prejudice to requesting parties and provides tangible benefits to producing parties who need to protect the privacy of their employees and customers, and the value of their commercial information. Historically, courts have permitted redactions of non-privileged material only in exceptional circumstances and generally limited them to privileged communications in otherwise responsive and non-privileged documents. Given heightened privacy and cybersecurity concerns, this culture must change.

In cases with personal information, data protected by data protection laws, or with documents that contain irrelevant sensitive commercial information permitting redactions should be the rule, not the exception.

To assist in this evolution, this article address seven misconceptions that have wrongfully limited the use of redactions.

Misconception 1: Requesting parties are entitled to irrelevant information in documents that contain responsive information. Much of the conversation around redactions begins with the incorrect presumption that because a document contains relevant information, the requesting party is entitled to the entire document. This is sometimes called the "completeness doctrine." However, this idea is not enshrined in the rules and, at best, is a custom built on the common-sense proposition that the parties need the context of relevant information to adequately understand the relevant information provided. That being said, requesting parties are not even entitled to all relevant information if it is not proportionate to the needs of the case. And, as Judge Moreno stated in the Takata Airbag MDL: "lt is only logical, then, that a party is similarly not entitled to receive every piece of irrelevant information in responsive documents if the producing party has a persuasive reason for why such information should be withheld." In re Takata Airbag Prod. Liab. Litig., 2016 WL 1460143 at *2 (S.D. Fla. March 1, 2016).

Misconception 2: The production of irrelevant information causes little harm to responding parties. This misconception is generally based on a pair of false premises. First, if the information is truly irrelevant then it will not matter to the litigation, so its production does not harm the responding party (i.e., it won't be used in the litigation and cannot prejudice the responding party in the case). Second, if there is a protective order, the information will not be disclosed to any third parties, so the minimal harm to the responding party is further reduced.

The first premise is myopic and the second ignores basic data security protocols. The mere unnecessary transfer and disclosure of irrelevant personal information invades the privacy of employees, vendors, consumers, and customers. These non-parties rarely have a "dog in the fight," but it is their personal information that is provided to requesting parties' counsel, experts, and in many cases, the requesting parties themselves. This prejudices these non-parties by exposing their personal information to persons they never knew would have it, and the producing party may want to protect these individuals from this unnecessary harm.

Protective orders are not a panacea. The more copies of information and the more distributed that information, the more likely it can (and will) be stolen, inadvertently disclosed and/or misused. It is well known that law firms and their agents are prime targets for cyber criminals and while receiving parties have an obligation to take reasonable steps to protect the information obtained in discovery, not all of them have the technical skills to meet such obligations and even where they do, they cannot guarantee the information's security, and usually balk at provisions (like indemnification language) that would mandate they do so. The safest and best way to protect irrelevant data is not to produce it in the first instance. With relevant and responsive data, it is necessary to take this risk, but with valuable or sensitive irrelevant information, redacting it from responsive documents is a sensible and reasonable precaution.

Misconception 3: Redactions undermine the context of the document. The most common argument against a more liberal use of redactions is that they degrade the requesting parties' ability to understand the document. This is not only easily solved, but ultimately a case of the tail wagging the dog. In the average complex litigation, the vast majority of produced documents are marginally relevant or redundant and are not used in any motions, hearings, or depositions. Thus, the value in understanding these documents is low. Moreover, for the documents where a requesting party believes that the redactions undermine their understanding, they can discuss this with producing party. For these limited documents, the producing party could lift the redactions or provide a quick peek of the document. In the unlikely event that the parties cannot agree about the redactions, these disputes can be raised with the court. However, the concerns about context are greatly exaggerated, as redactions on many files do not raise any such concerns (e.g., irrelevant entries on spreadsheets or irrelevant personal information on reports (SSN)).

Misconception 4: Redactions will be used to hide important relevant information. Another common complaint against redactions is that they will be misused because producing parties will redact relevant and responsive information to hide it from requesting parties. First, while all discovery rules can be abused, the rules of ethics and civil procedure are premised on parties and their lawyers acting in good faith. In fact, given that we trust lawyers and parties to produce entire documents that are harmful (in which case, unlike redactions, there would be no direct evidence that the documents had been wrongfully withheld), it seems that redactions are safer from abuse because the requesting party can at least see the redaction in the document, and challenge any redactions that are unclear on their face. In addition, a producing party might offer a requesting party a quick peek at a discrete number of redactions (so long as it is not abused) that could assure the requesting party of the legitimate nature of the redactions.

Misconception 5: Redactions will be over-used. In pushing back on redactions, requesting parties often allege that they will be overused and they will receive documents that are more blank spaces than they are content. First, this ignores the fact that if a responding party over uses redactions, it will draw the ire of the court and risk the judge ordering the party to lift all of its redactions, exposing information that is truly sensitive and irrelevant and wasting all the money the party has spent redacting. More importantly, redactions are self-policing as redactions are one of, if not the most, expensive element of discovery. Redactions are very time consuming and can make the review of documents 30-50 times more expensive. This is a built-in incentive to only redact only the data that is truly sensitive or private.

Misconception 6: Documents largely address one topic. One of the hidden assumptions that is the foundation of the other misconceptions is that documents largely address one issue and that in order to understand the responsive portions, one need to be able to review the entire document. While this assumption may be true in some cases, it is rapidly becoming obsolete in the information age where electronic communications and dynamic, interactive documents aggregate custom information to individuals based on their own preferences and historical conduct. Many reports discuss multiple different products only some of which are relevant to the litigation. Emails and text messages intersperse personal tidbits with responsive business discussions. Social media and user interfaces at work pull data from multiple databases and locations to present information on a multitude of topics (and, in the latter example, for employees to perform their jobs).

Misconception 7: Since they are withholding information, the redacting party should serve a "redaction log." This misconception creates a false parallel between withholding relevant documents that contain privileged information and redacting irrelevant information from otherwise produced information. While it is true that in both cases the producing party is withholding information, this is where the similarities end. When a party redacts a document, the receiving party knows it has been redacted because it can see the redaction (ideally, the producing party should include a tag for this in the metadata when it is produced so that it can be easily searched). A privilege log is necessary because the party needs to explicitly make the claim of privilege, but more importantly, without the log the requesting party would never know the document was withheld and the requesting party could not test the assertions. Moreover, the privilege log provides basic contextual information about the document (e.g., date, author, recipients) that is self-evident with a redacted document. So long as the producing party provides the reason for the redaction (e.g., irrelevant sensitive information or personal information or data protection), there is no reason to undertake the expense of creating a log that contains the same information.

* * *

Because of heightened privacy and data security risks, the time has come to change the culture around the use of redactions in civil discovery: The fears around redactions are exaggerated; the likelihood they will be overused is not significant because redacting documents is so expensive; and blatant misuse is easily spotted and corrected. Permission to redact irrelevant personal and commercially sensitive information should stop being the exception and become the rule.

David Kessler is head of data and information risk, United States, at Norton Rose Fulbright US. Andrea D'Ambra is head of e-discovery and information governance, United States.