Audio Discovery – The Next Frontier

June 1, 2008

From March 2009 UK financial institutions will have to record and store telephone conversations and electronic communications relating to client orders, under new regulations introduced by the FSA.  Under these new rules the information gathered will be retained on file for six months, in an attempt to combat market abuse, particularly insider dealing and market manipulation.

Until recently, inadequate technology solutions, a lack of notable common-law decisions and a general reluctance among litigators has kept the exchange of audio files in legal discovery from becoming commonplace.  However, recent rule amendments, developing case law and burgeoning technology are steering practitioners in a new direction. The need for e-discovery has surged and electronic data has had an impact in a vastly increased number of cases.

Audio evidence can be compelling, in the context of settlement discussions and at trial, especially if timing, message content and vocal intonation play a substantial role.  When a sound recording is played for a jury, the jury not only hears the witness’ words, but the tone, expression and other subtle clues inherent in speech.  Further, sound recordings can be an especially rich evidentiary trove as some individuals may intentionally choose to leave “smoking gun” information in a voicemail instead of an e-mail, given the fact that e-mail mining has become standard course in many cases.

“Sound recordings” are now defined as a discoverable type of electronically stored information (ESI), that is subject to the same preservation and production requirements as other documents and ESI.  In light of this, litigation teams will need to consider methods for collecting, reviewing, and producing audio files in a reasonably usable format.

Technical Developments

Audio discovery technology is now addressing some of the limitations with listening and transcription review. Legal teams are no longer required to listen to hours and hours of audio files or have the entire data set transcribed.  Instead, audio discovery technology makes sound recordings easily searchable in a quick and cost-efficient manner.  It does this by breaking the audio data down into phonemes – the smallest components of human speech. After processing, the audio files are uploaded into online review software. 

Reviewers can use online software to search the hours of audio for specific words and phrases, jumping to targeted portions of the audio file for listening purposes.  Within the online software, responsive documents can be categorised and the audio discovery service provider can provide the legal team with native audio files.  The technology also works for audio portions of video files. 

The technology to digitally record speech is progressing rapidly, making audio files more like e-mail and file data every day.  For example, today’s computerised phone systems and voice recording systems typically store voicemail recordings as digital sound files.  For quality control purposes, businesses are also seeing the benefit of recording voice discussions digitally in corporate call centres. 

The most common sound file formats are: .aif, .aud, .mp4,.vox, .aifc, .avi, .mpeg, .wav, .aiff, .mov, .mpg, .wma, .asf, .mp2, .mpg4, .wmv, .au, .mp3, .qt. These files can be saved, copied and preserved like other digital data files.  Some sound archiving technologies can also add metadata to audio files, such as date and time stamps; recipient and sender telephone numbers; identification of actions taken in a telephone system such as “retain,” “forward” or “delete.”  Like other forms of electronic evidence, because the audio files are stored digitally, the volume of these types of files within a corporation is increasing massively. 

In addition, e-mail and voicemail systems are converging.  E-mail is now accessible via telephone voicemail systems, and voicemail is now being sent to recipients by e-mail.  Voicemail systems used by businesses today send and store voicemail through a central computer system which converts the audio sounds into digital bytes.    Accordingly, what began as an audio voicemail message may end up as an e-mail attachment.  Businesses are quickly moving towards these computer-based messaging systems, called Unified Messaging Systems (“UMS”). 

Lastly, because most phone and sound recording systems are based on some version of a standard computer operating system, it is possible for a corporate information technology department or electronic evidence service provider to harvest raw data from a modern sound recording system.  The prudent path for any business or lawyer is to handle voicemail like traditional e-mail and other discoverable ESI. 

Audio Discovery Options – searching, review and production

As with traditional discovery, lawyers must collect, search, organise and review audio files before producing the data to the opposing party.  In the past, parties that were compelled to produce audio files ordinarily reviewed them by either listening to each file, or sending the files to a transcription service and subsequently reading the transcribed text.  These methods are arduous and slow, complicating the review of small volumes of recordings and making large volume reviews virtually impossible.  For instance, listening to sound files during review can take up to three or four times the actual duration of the recording.  Another limitation associated with listening to audio files is the lack of searchability.  Without search functionality, legal teams must rely on the reviewers’ abilities to identify keywords as they listen to the recordings.  On the other hand, if the recordings are transcribed to text, the legal team must wait weeks or even months before review can even begin.  Another downside to transcription review is that any tone or inflection in the recording will be lost, without going back to the original sound files. 

Audio Discovery Best Practices

The discovery of audio files is quickly becoming the next frontier in legal discovery.  Below are five steps for your next audio discovery project.

1. Survey your client’s sound recording systems
2. Discuss sound recordings with the opposing party and court
3. Develop specific discovery requests; object to overbroad requests
4. Review and produce audio files
5. Consult an audio discovery service provider

Alexandra Harrison is a Legal Consultant at Kroll Ontrack, experts in computer forensics & electronic disclosure: