Courts are routinely rejecting discovery demands for “all” social media evidence. Such demands are understandable – the volume and type of materials posted to social sites is expanding and users are increasingly protecting content from public eyes with more comprehensive privacy controls. Yet, courts are holding that litigants have an interest in keeping irrelevant information confidential and that reviewing such information in discovery (possibly several social media accounts spanning years) can be burdensome on both parties and courts. To secure an order for the discovery of private social media information, requesting parties often need to make a predicate showing that there is a likelihood that the information behind a social media privacy “wall” is relevant.
But, if the information is behind a privacy wall, how does a litigant make the showing? Requesting parties have relied upon “publicly available” materials (such as a profile picture or blog post), depositions of the account holder or third-party information to make the showing. Yet this procedure does little to address the burden of reviewing potentially voluminous, irrelevant material (statistics vary, but the average Facebook user has over 200 photos, likes 8 things a day, has over 100 friends, etc.) or address when an in camera review is appropriate.
A suggestion: Seek less
Rather than engage in this uncertain and potentially time-consuming process, litigants may wish to consider making narrowly-tailored, specific demands. In fact, two recent decisions highlight that some claims do not require access to social media content at all. In these cases, evidence of the act of using social media (or text messaging) may be all that is required. In these situations, privacy concerns may not be as heavily implicated and a court may be more likely to allow disclosure.
Definitions & privacy
Before beginning let’s define terms. For our purposes, Content is the message a social media user wants to communicate to others, such as: “I don’t understand the appeal of Justin Bieber.” Common social media content might be Facebook wall posts or “likes,” tweets, pictures, etc. Non-content is information related to the act of using social media such as information showing a message was sent, received, read or edited, etc. Content and non-content will often contain a time-stamp indicating when a user performed a particular act and what kind of device he or she used.
Why the distinction? Context matters in privacy arguments. Even if you seek non-content, the opposing party may raise an objection that the request invades recognized privacy expectations. In Detraglia v. Grant, 68 A.D.3d 1307 (NY AD 3d. 2009), the plaintiffs sought cellphone and wireless data records for devices in the vehicle of a defendant who struck the plaintiffs’ car. The plaintiffs also sought to depose the IT specialist responsible for the devices at issue.
The court recognized that users have an expectation of privacy in telephone records listing telephone numbers. In response, the plaintiffs produced evidence that the defendant may have been distracted immediately before the accident and submitted the affidavit of a tow truck driver stating that he saw a laptop computer open and running in the defendant’s vehicle shortly after the accident. This showing was sufficient to suggest the records may be relevant.
The appellate court ordered an in camera review of the records to determine if relevant evidence existed and to protect privacy, see also Dietrich v. Buy-Rite Liquidators, Inc. (2012 PA Dist. & Cnty. Dec. LEXIS 158)(citing the privacy of telephone records and referencing case law regarding social media discovery, a court refused to allow the discovery of cellphone records where the defendant did not produce any evidence that the plaintiff was using his cellphone at the time of the accident. The court also refused to undertake an in camera review of the records citing the burden on the court).
Merely sending texts may be the basis for liability
In the personal injury action Kubert v. Best, 2013 WL 4512313 (NJ App. Aug. 27), a New Jersey appellate court found that a defendant who SENT a text message to a driver could be liable to plaintiffs struck by the recipient driver. The sender could be liable if he or she knew or had special reason to know that the recipient driver would read the message while driving.
Here, the court did not have the text message content before it. Although the court acknowledged that message content would be useful, it went on to analyze the relationship and communications patterns of the defendants and the time-stamps of the text messages in relation to the time of the accident. The Kubert court found that summary judgment in favor of the sending defendant was appropriate because the plaintiffs had not presented evidence that she knew or had reason to know that the recipient driver would look at the message while driving.
Building on this case, one can readily imagine scenarios where a sender would be likely to know that the recipient will read the message while driving, such as an ex-spouse texting a driver about dropping off children while that driver is en route or a supervisor demanding an immediate answer from an employee who routinely drives long distances as part of a job.
Using posts to establish whether employees received meal breaks
Another example of using non-content to pursue a claim is a recent decision in the wage and hour action Jewell v. Aaron’s, Inc., No. 1:12cv0563, U.S. Dist. LEXIS 102182 (N.D.GA July 19). In this class action, the plaintiffs alleged that they were not given meal periods. The defendant received an anonymous tip that the named plaintiff, Kurtis Jewell, frequently posted to social media during his alleged meal time and the defendant had apparently reviewed some content from Jewell’s Facebook account.
The defendant made a broad request for “all documents, statements, or any activity” posted on the Web or on social media sites. The request was directed at a sample of the claimant pool: 87 of the plaintiffs. The defendant’s request was also limited in time. It sought posts made “during working hours” over a 4-year period.
This limitation added an additional difficulty for the plaintiffs as individual meal periods were not at a set time (a “potentially moving target” in the words of the court). To support its request for the information, defendant asserted that such posts may show that a sample plaintiff posted information about meal periods or engaged in a successive pattern of posts indicative of having a meal period.
The plaintiffs responded that whether an individual plaintiff used social media was irrelevant to whether he or she received a legitimate meal break and that compilation of the information the defendant sought would take from 1,323 to 26,462 hours. The court rejected the defendant’s request. It found that the use of social media “may have no bearing” on whether a plaintiff received a meal period, that the defendant’s factual predicate was purely speculative, and that the request was unduly burdensome.
The Jewell defendant may have had a stronger argument if it only requested evidence of when posts were made (e.g. posts made during possible meal periods), thereby reducing the material to be reviewed and not seeking content the plaintiffs deemed private. In fact, as the vast majority of Twitter users’ tweets are publicly available, the defendant could have requested the Twitter identities (aka “handles” such as @ScottMalouf) of the 87 plaintiffs, if any, and then performed its own analysis of when posts occurred.
Assuming it found a sufficient pattern of usage during alleged/likely meal periods, the defendant could have used that information in support of a predicate showing.
Litigants should keep the following in mind when seeking discovery of social media:
1. Social media demands should be narrowly-tailored to discover relevant information.
2. Understanding the kinds of information available on a social media site is necessary to assessing privacy concerns, burdens and relevance.
3. Relevant information may be on a “public” site like Twitter.
4. Non-content may be sufficient to prove a claim or as an initial discovery step.
5. When seeking discovery of material behind a privacy wall, the litigant may need to make a predicate showing that the material sought is likely to be relevant.
Less (may be) best.
Scott Malouf is an attorney who helps other attorneys use social media, text and web-based evidence. You can learn more about him at his website (www.scottmalouf.com) and follow him on Twitter at @ScottMalouf. Although he is a good cook, he has been accused of “over herbing.”