Can Voice Recognition Software Ever Replace Human Stenographers?

In an effort to cut spending, many courts are attempting to replace certified court reporters with digital recording devices, or a combination thereof.  By cutting the expense of hiring a stenographer, the courts have to consider the sunk costs of transitioning their courtrooms and buildings over to a digital recording system.  But the incorrect perceptions regarding the advantages of digital recording systems don’t stop as labor costs.  There are several other considerations that have proven to be false.


Digital recording systems sometimes seem to be less expense – the thought being that it’s a one-time purchase that will help alleviate the costs of human salaries and error.  However, a digital recording system has to be purchased, hard-wired, and like all technology, updated every few years.  The initial purchase of the equipment ranges in the mid $2,000 range, while hard-wiring a courtroom can cost up to $20,000.  Network upgrades to store the large digital files can cost upwards of $50,000.

Court reporters are often contracted or freelance employees.  They carry the responsibility of managing and paying for their own equipment, alleviating municipalities of those costs.  Though digital recordings themselves do not demand an hourly wage, they do need to be managed by a paid monitor or manager. The need for human management over digital recording devices can often be complex.  Think of a dedicated IT professional whose job involves a myriad of issues involving the wiring, audio, software, and data storage issues that come along with switching over to a digital system.

Beyond equipment costs, court reporters are paid per case, as well as per page of transcript.  According to national reports, the average court reporter’s compensation per page is far lees that the cost to transcribe a digital recording.  Some California courts are paying a reported $9 per digitally transcribed page.  The supposition that alleviating stenographers will save the courts money is false.  Though staffing requires less spending, hard-wiring courtrooms and purchasing the necessary technology is a hefty up-front expense, and the cost of transcription of digital recordings is typically equal to or much more expensive than that of hiring a court stenographer.


Cost is not the only perceived upside – increased accuracy is also touted as a feature of digital court reporting systems.  However, this perceived benefit is also usually laden with potentially costly landmines that can result in increased litigation costs and even mistrials.  Automated software simply does not have the ability to react to the emotions and contentious dialogue that occurs during court proceedings.  Digital recordings cannot stop proceedings or separate out noises from each other if mumbling, sobbing or heavy accents make a testimony inaudible.  Further, digital recording equipment cannot discern between what is on the record and what is off the record.

Here are a few examples:

  • Digital recording machines have malfunctioned, failing to record testimonies and have caused costly and lengthy mistrials (in 2007, an Essex County, NJ digital recording system failed to record testimony, resulting in a mistrial of a Medical Malpractice lawsuit).
  • There have been instances where digital recorders overhear attorney/client conversations, sometimes making privileged information official record.
  • There have been instances where testimony has been lost because the equipment cannot discern what is being said over crying, laughing, and other individuals speaking at the same time.

In high impact cases, as well as cases with heavy transcription needs, there is simply no substitute for the presence of a live stenographer.  Many courts that have made the switch to digital systems citing accuracy and cost concerns are now finding it more costly than originally anticipated.  The costs to install and maintain the technology and additionally, repeat proceedings where audio is inaudible or inaccurate, and in a worst-case-scenario, retry a case, are proving to not be worth it.


A study done by the California Court Reporters Association

About the author:

Lance Brusilow is the owner of a Philadelphia court reporting agency called Brusilow+Associates. He has been serving the Philadelphia area for over 30 years.

Register for Updates

Register to receive the latest news and updates when we post. We regularly publish no-fluff articles on our website. The articles range from topics that specifically apply to attorneys, paralegals, legal assistants, or court reporters, to topics that apply to most legal professionals.

Thank you for registering for email updates!

2 replies
  1. Noah Kovacs says:

    I do not see this being a replacement by any means. There is a major gap for machine error. Perhaps keeping these as a back up or contingency plan maybe but not a full replacement. Thanks for sharing. Pretty interesting.

  2. Julia says:

    One question I have is about the use of Voice Writers. At this point in time voice recognition software can recognize with ease speeds of 180 to 200 wpm when dictated by reporters who use the Stenomask. Stenomask reporters can dictate speeds far in excess of what the voice recognition software can recognize. How long will it take voice recognition software to advance to where it can recognize 250 to 300 wpm? Eventually, it is the voice writers who may be competing for court reporting jobs. I am not worried about electronic court reporting as I know it is not accurate.

Comments are closed.