A series of trades engaged in by a financial services company employee have become the subject of regulatory interest and private litigation. The company’s primary regulator has requested that the company “[p]roduce copies of all recordings and or audio files of conversations or telephone calls related to or concerning the transactions at issue.” The private plaintiffs’ lawyers have made a similar request, though it does not specifically seek “audio files.” Rather the private plaintiffs’ request seeks “all documents related to or concerning the transactions at issue.” The company is considering how to respond, and whether it can avoid the costly and burdensome task of collecting, processing, reviewing and producing recorded audio files.

Must Audio Recordings Be Produced?

Under Federal Rule of Civil Procedure 34, electronically stored information (ESI) is discoverable if it is “stored in any medium” from which it can be obtained “either directly or, if necessary, after translation … into a reasonably usable form.” To the extent that audio recordings are created and retained in the ordinary course of business, they may, like email, constitute discoverable ESI.

On the other hand, Federal Rule of Civil Procedure 26 protects a party from being forced to produce ESI from sources that that are “not reasonably accessible because of undue burden or cost.” There has been little explicit guidance from the courts on the degree to which this rule may apply to requests for large amounts of audio data. However, in other ESI contexts, courts have crafted novel approaches to balance discovery rights against burden and cost concerns rather than preclude discovery entirely. For example, although courts typically rule that requests for production from email backup tapes impose an undue burden, some courts have allowed restoration of a limited set of backup tapes, at shared or shifted cost, where emails from the relevant period have not been retained.

In the regulatory context, the creation, retention and production of audio recordings is compulsory under certain circumstances. The Securities and Exchange Commission (SEC), the Commodity Futures Trading Commission (CFTC) and certain other regulatory entities require that certain oral communications—e.g., telephone and “squawk box” conversations—be recorded and maintained for certain individuals—e.g., Broker Dealers (SEC) or Registrants (CFTC)—for a certain period of time. This requirement permits the regulator to demand and review such data during the course of an investigation.

Sources of Audio Data

There are several potential sources of audio data, including:

  • Voicemail. Voicemail is the most common source of audio data. Potential sources include work or personal cell phones or land lines.
  • Recorded telephone lines. Telephone lines are most often recorded because it is either required for regulatory purposes (as discussed above) or desired for business purposes. For instance, a company’s call center may record customer service calls for training and quality control purposes.
  • Recorded squawk boxes. “Squawk boxes” are intercom systems used by broker-dealers and other financial professionals to broadcast offer, bid and other market information to traders and other market participants. Like recorded telephone lines, if a Broker Dealer, Registrant or other regulated individual uses a squawk box for voice communication, those too are likely subject to record keeping requirements.

Strategies and Best Practices for Dealing with Audio Data

As a practical matter, regulated entities often simply comply with a regulator’s requests because the requests concern calls made only by one or two individuals over a short period of time (e.g., recorded calls over the course of a few days). In the civil context, however, the requests are likely, at least initially, to be much broader and cover more custodians over a longer period of time.

Audio reviews, like written document reviews, can be costly and time consuming: vendor and reviewer costs can be significant. However, there may be ways to attempt to narrow such broad requests that are likely to satisfy the regulator or the judge and help keep the costs and burden to a minimum.

  • If at all possible, review and produce written communications before audio data is considered. This allows both parties to learn more about the case so that a targeted audio review protocol can be developed and agreed upon. The written communications will often allow the parties to narrow the individuals and dates of interest. Also, the written communication review will educate the parties on which search terms are likely to yield the correct results in an audio review.
  • Squawk boxes typically record 24 hours a day, seven days a week. One easy way to limit the amount of squawk box data to review is to seek agreement to process only data recorded during work hours.
  • Some e-discovery vendors have the capability to process and search audio data. In processing audio data, vendors often convert speech to dictionary-recognized text. Any word for which no dictionary entry match is found is rendered phonetically. All text (including phonetic spellings) is indexed. Once the audio data is converted to text, standard boolean text search operators, such as strings and wildcards, can be used to identify potentially relevant results. However, as with written documents, it is possible that not all audio data will be rendered to text correctly. This is especially true with jargon or when a speaker has a heavy accent or switches between multiple languages in a single conversation. As such, it may be necessary to “tune” the rendering tool and the search terms multiple times throughout the process. Once rendered, the text is then paired with the audio, which is streamed through a review tool with the “hits” highlighted as a visual representation of the waveform. The reviewer is able to review the audio where the hits occurred.

In conclusion, when possible, it is helpful to anticipate discovery concerns and develop a plan to mitigate the risks that arise in preserving and producing audio data. Counsel can consider instituting a plan for collecting audio data, and whether the litigation hold letters specifically identify audio data sources on company-issued and personal devices as among the categories of documents and data to be retained, both to ensure that the clear instruction is given and to document compliance with preservation obligations. Counsel also should understand the capabilities of the company’s preferred e-discovery vendors and if they do not have the ability to process and host audio data, such a vendor should be identified.