By Tom Fernicola

If you want to understand how a profession dismantles itself, don’t look at the people attacking it from the outside. Look at what the people inside were arguing about while the gates were left wide open.

For the better part of two decades, the stenographic profession has been locked in a bitter, circular debate over theory. We argued about stroke-intensive writing versus brief-heavy systems. We elevated fast writers to the status of gurus. We argued about which method was to blame for a catastrophic student dropout rate that has hovered above 90 percent for as long as anyone has been counting.

And while we were having those debates, a billion dollars in venture capital mobilized to replace us with AI legal transcription.

The tragedy of the court reporting profession is not that we were defeated by superior technology. The tragedy is that we handed the technology companies the exact excuse they needed to replace us, and we manufactured it ourselves.

The Manufactured Court Reporter Shortage

The current court reporter shortage is not an act of God. It is not a demographic inevitability. It is a manufactured crisis — the direct result of a profession that refused to fix its own educational pipeline.

When a profession fails to graduate 90 percent of its students year after year, decade after decade, a shortage is a mathematical certainty. Instead of confronting that structural failure honestly, the profession looked for saviors. It found educators who promised that their specific theory would “revolutionize the profession in the direction of faster graduation rates.”

The profession believed them. The National Court Reporters Association endorsed them. And the graduation rates did not move.

But the failure to graduate students did something far more damaging than just shrink the workforce. It created a vacuum in the courtroom. When a judge cannot find a stenographer to cover a docket, they do not cancel the docket. They look for an alternative.

The digital court reporting and AI transcription industry did not create the shortage. They simply walked into the vacuum that our theory debates and educational failures left behind. We built the crisis; they built the business model to exploit it.

The Speed God Failed

Perhaps the most fatal mistake the profession made during those decades of internal debate was what it chose to worship.

We built a credibility hierarchy around speed. We held contests. We handed out trophies. We treated speed champions as the ultimate authorities on how to train, what to teach, and what the profession should value. Speed became the credential that mattered above all else.

It is the equivalent of a profession of ditch-diggers spending thirty years arguing over who has the fastest shovel, handing out gold shovels to the winners, and making the fastest shoveler the head of the academy — all while John Deere is parked at the edge of the site with an excavator.

Those ditch-diggers are now grave diggers, and the graves they are digging are their own.

You do not compete with a machine on the machine’s terms. It is a losing proposition to attempt to overcome one’s biology.

Aside from one person who can write 360 words per minute for a single minute under controlled conditions, human speed is pathetic compared to what a machine can do.

We cannot compete with machines on processing speed. A server farm does not get tired. It does not have a bad day. It processes speech at volumes and velocities no human can approach, around the clock, at a fraction of the cost.

By deifying speed, the profession built its entire credibility structure around the one attribute that technology renders completely irrelevant. When we needed to make the case for our own survival, the argument we had spent thirty years building was “we are fast.”

The machine answered: not fast enough.

The only competitive advantage a human stenographer actually has over an AI is the ability to handle novel, unpredictable speech in real time with certified, accountable accuracy. But instead of defending that advantage — instead of demanding rigorous, public data on error rates and human accountability — we were busy timing speed runs.

The Billion-Dollar AI Transcription Incentive

While we were polishing trophies, the money moved in.

The global AI voice transcription market is projected to reach $42 billion by 2030 [1]. In early 2026, Tyler Technologies — a $20 billion software giant that runs the case management systems for hundreds of U.S. courts — paid $212.5 million to acquire For The Record, a digital courtroom recording company [2]. Steno, a “tech-enabled” litigation provider, raised $49 million specifically to push its AI transcript tools into the country’s largest law firms [3]. Verbit, an AI platform explicitly targeting the legal sector, raised over $550 million at a $2 billion valuation [4].

You do not deploy a billion dollars to capture a niche. You deploy it to capture an industry.

They will market their systems as faster, cheaper, and “good enough.” But they have a structural problem: an algorithm cannot swear an oath. An algorithm cannot be held legally accountable for the accuracy of a record. The legal system still requires a human being to sign the certification page.

And that requirement is creating a new, deeply cynical role for the very professionals the machines are displacing.

The Last Job in Court Reporting

What does the endgame look like for a displaced court reporter?

They will not be producing the record. The machine will do that. But the agency selling the machine’s output still needs a credentialed professional to take legal responsibility for it.

So the agency will offer the displaced reporter a job: reviewing and certifying AI-generated transcripts.

And they are giving those roles titles designed to keep the profession from figuring out what is really happening.

In January 2026, Planet Depos — a global agency that explicitly markets its “integration of artificial intelligence” — announced it had hired Kristina Tan, CCR, as its new “Director of Reporting Technology.” [5] The announcement was framed entirely as an elevation of the profession. Her role is to “bridge the gap between reporters and rapidly evolving technological advancements” and oversee an internal scoping team.

The title sounds like empowerment. The reality is displacement. Putting a credentialed reporter’s face on the AI integration process gives the agency a human shield against the argument that they are replacing reporters. It signals to the reporter community that embracing this transition is a career advancement rather than a capitulation.

But for every one “Director of Reporting Technology” role created to manage the AI infrastructure, that same infrastructure eliminates the need for dozens of freelance reporters. The math is not in the profession’s favor, and the titles are designed to obscure the math.

But look closely at how these jobs are structured. The reviewer will not be paid by the page, because paying by the page gives the reviewer a financial incentive to slow down and actually read carefully. Instead, they will be made salaried employees, or paid a flat per-transcript fee, to process volume.

The economics of the AI model demand throughput. The reviewer will be required to certify hundreds, possibly thousands, of pages a day — far more than any human could genuinely read and verify against the audio. They will be reduced to spot-checking, trusting the machine’s confidence scores, and signing their name to work they did not produce.

Their income will be dramatically reduced. They will lose the professional premium of being the sole producer of the record. They will lose their independence. They will be reduced to cheap labor, sitting at a desk, clicking through AI-generated transcripts, putting their license and reputation on the line for an agency that is billing the client full price and keeping the margin.

They will become a cog in the machine that replaced them.

This is the ultimate irony of the theory debates. The brief-heavy systems that were supposed to save the profession actually trained reporters to do exactly what AI does best: rely on memorized coverage of predictable, high-frequency phrases. When you train a human to act like a machine, you make them perfectly replaceable by one.

We argued about how to write the record, while the industry outside the room built a billion-dollar machine to take the record away.

And what is the profession’s response?

Complain on Facebook. Argue about theory. Post memes about digital reporters.

But individual inertia is only half the story. The true betrayal came from the top.

The National Court Reporters Association — the organization that collects our dues, runs our PAC, and claims to protect our future — saw this coming. They had the resources to fight it. They had the institutional relationships to organize a defense. Instead, they spent the last decade platforming the very people and narratives that kept reporters distracted while the industry reorganized around us. They handed out trophies while the gates were left wide open.

The AI companies, the deposition agencies, and the corporate investors are not afraid of our forums. They are counting on them. They know that a profession that spends its energy arguing with itself — led by an association that refuses to fight the actual war — is a profession that will not organize to stop them.

They bet on inertia — they’ve got it — and they bet on our leadership’s complicity.

Looking at the scoreboard, they won both bets.

Soon, the only job left for a court reporter will be signing the machine’s work.

References

[1] Strategic Market Research. “Speech and Voice Recognition Market.” 2024.
[2] GovTech. “Tyler Acquires Court Tech Firm For the Record for $212.5M.” February 3, 2026.
[3] Steno. “Steno Secures $49M Series C to Fuel Rapid Expansion.” March 26, 2026.
[4] TechCrunch. “Verbit lands $250M Series E at a $2B valuation.” November 23, 2021.
[5] Planet Depos. “Kristina Tan, CCR, Joins Planet Depos as Director of Reporting Technology.” PRWeb, January 26, 2026.