Information about Information: The Increasing Role of Metadata in Bankruptcy

Bankruptcy courts are faced with increasing discovery disputes over electronic discovery. One increasingly prevalent topic of contention is the production and use of metadata, which can be used in a variety of ways in avoidance actions and other adversary proceedings. This article will outline what metadata actually means, how it is produced, and some of the issues that can arise both inside and outside of bankruptcy in the discovery and use of metadata.

The dictionary definition of metadata is, simply, “data about data.”[1] In litigation discovery, metadata is best understood as information about a document other than the content of the document visible in a printed copy of the document.

The types and amounts of metadata available for a particular document being produced in discovery vary based on the original format of the document. For example, an electronic email file such as .msg file or .html file (as opposed to a printout of the same email) would include, among other metadata, date sent, time sent, to, from, cc and bcc. A Microsoft Word file, on the other hand, even if it was a memorandum sent to the same individuals as the email, would not include the sender and recipients as metadata, but might include other information such as the date created and date last modified.

How Metadata Is Produced

Generally, metadata is commonly produced in two ways. A party can produce responsive documents by providing native files — copies of the original electronic file for the responsive document in the format used by the application that created the document. By providing the electronic file, a producing party is providing the metadata contained in that file.[2] The other common approach to producing metadata is for a party to provide bates-numbered images (a static version of the document branded with bates numbers and confidentiality designations) accompanied by a load file that provides metadata for each document. In this type of production, the producing party determines what types of metadata it will produce and provides those metadata fields, identifying the bates-numbered document.

One primary advantage of native production is the simplicity of the production process. Native documents can be produced by simply copying the data to a production media, which requires less technical expertise than the creation of a load file and bates-labeling production images. There are disadvantages to this approach, however. One practical challenge is that producing documents in native means that the parties do not have a shared, bates-labeled document to reference as an exhibit in pleadings, depositions and trial. Printing documents with different software and specifications may result in different, or even misleading, versions of the same document, and relying on an opposing party to label documents correctly can be an issue. Also, unlike bates-labeled images, native documents can be modified, sometimes inadvertently. And native documents also cannot be redacted without changing document metadata.

Providing bates-numbered images and a load file provides both parties a shared evidentiary copy and gives the producing party more control over the specific metadata it produces. However, it generally requires a review tool and a vendor or internal specialist to run the production using the tool and generate the load file. The process of generating the bates-numbered images also has more challenges than copying native files, and more opportunities for controversy over what was and was intended to be produced.

Considerations as a Producing Party

If an opposing party is seeking metadata, you should first clarify the specific information they are seeking. Often requests like “all metadata” or “native files” seem straightforward to a requesting party, but they create complications from a responding perspective. For example, email files may be collected in a .pst container format for review, then converted to html to be reviewed as individual documents, remaining “native documents” but necessarily in a different format from the one in which they were collected. While .pst would technically be a “near-native” format, some parties would expect a requested native production to be in .pst format, while others might claim it did not qualify as “native.” If you are producing metadata, you also need to be fully aware of the metadata being produced, which requires reviewing the metadata to determine whether it contains privileged information. As courts have cautioned, “Metadata ... may contain privileged or protected information.”[3] Thus, a producing party should review document metadata being produced the same way it would review any other document content.

Considerations as a Requesting Party

Metadata can be a useful tool in the bankruptcy context. For example, for a party pursuing an avoidance action, such as recovery of a preferential transfer under § 547 of the Bankruptcy Code, metadata may be able to tell you when an invoice was created, when it was sent to the debtor, and to whom the invoice was sent. Metadata may also be able to assist a party in identifying when a transfer occurred or when a party made a demand for payment. In addition, metadata may help identify attempts to cover up fraudulent transfers by showing when documents were edited and, in some cases, what edits were made.

Pursuant to Fed. Rule Civ. P. 34(b)(1)(C), a requesting party “may specify the form or forms in which information is to be produced.” This permits a party to seek metadata as part of its requests, and a court will not look favorably upon a party seeking metadata or native production that did not so specify in its document requests.[4] Another consideration is reciprocity. Generally, it is inadvisable to ask for metadata that you are unwilling or unable to provide in response to the opposing party’s requests.

As bankruptcy courts become more active in reviewing and rendering decisions on discovery disputes, it is critical that bankruptcy practitioners understand the concept of metadata and the role that production of metadata can play in their causes of action.