The displacement of stenographers by digital technology was not caused by artificial intelligence, but was made possible by a self-inflicted labor shortage driven by a 90 percent dropout rate in steno schools.

A major national court reporting agency recently created a new role: Director of Reporting Technology.

The announcement was framed as support for you — better tools, faster workflows, improved integration between reporters and evolving software. The framing was professional and well-intentioned.

It was also a signal you should not miss.

A Director of Reporting Technology is not a support role. It is a systems role. And once a company organizes around systems, it stops building around people.

The infrastructure taking shape at agencies across the country — AI-assisted transcription, internal scoping teams, fast-turnaround certified transcripts on digital proceedings — functions to future-proof the company, regardless of how it is framed.

Protecting your position and future-proofing the business are sequential goals, not the same goal. The agencies need you now. The systems they are building are designed so that need diminishes over time.

The important thing to understand is this: the agencies did not create this situation. They are responding to one that was handed to them — by the profession itself.

To see how completely the loop has been closed, look past the title to the operational details that rarely make headlines.

Some agencies have been cross-training their videographers as digital reporters. Think about what that means. The videographer is already in the room. Already on the agency’s payroll. Already recording multi-channel audio. Cross-training that person as a digital reporter costs relatively little — but it means the agency now has a dual-function employee who can cover a deposition without scheduling a freelance stenographer at all:

One person. One fee. Zero external dependency at the capture layer.

Pair that with an internal scoping team — not freelance scopists the reporter hires and pays directly, but agency employees supervised by the Director of Reporting Technology — and the picture becomes exact.

Their employee captures the audio. Their AI produces the draft. Their scopists clean it. Their platform delivers it.

It works. And it works without you.

You are not required for that loop to function. Your presence becomes a preference — not a dependency. The freelance stenographer is becoming an optional add-on for clients who specifically request one — not the default infrastructure of the proceeding.

The in-house scoping team is the detail most reporters have not focused on, and it may be the most consequential move of all. Freelance scopists have historically been an independent layer — reporters hired them, paid them directly, maintained those relationships outside agency control.

When an agency builds an internal scoping team, it is not trimming costs at the margin. It is severing the last external dependency in the transcript production chain. The agency now owns the entire workflow from room to delivery.

This did not happen suddenly. The announcement of a Director of Reporting Technology was not the beginning. It was the public acknowledgment of something already operational — the formal title placed on infrastructure that had been assembled quietly, piece by piece, while the profession was focused elsewhere.

The shortage came first. AI filled the vacuum.

Across multiple industry reports and state licensing data, the pattern is consistent. The numbers establish what happened before AI entered the picture.

Over the last decade, the stenographic workforce dropped 21%. Approximately 23,000 reporters remain nationwide, and industry projections suggest the workforce could fall below 19,000 by 2029. Enrollment in steno programs has collapsed 74%. Nearly half of all programs have closed.

California — the largest legal market in the country — issued 121 new individual licenses in 2023-2024, against a need for hundreds just to meet minimum court requirements. The average NCRA member is 56 years old. Half the employed workforce was already retirement-eligible as of late 2023.

These are not technology statistics. They are training statistics.

The pipeline collapse happened before AI was capable of replacing a reporter. It was not caused by ChatGPT or automated transcription or agency consolidation. It was caused by a 90% dropout rate in court reporting schools that the profession never fixed — and, for structural reasons, never had a strong incentive to fix.

That 90% dropout rate is not a student failure. It is a design failure. A 90% failure rate is not a filter. It is a signal the system is broken. Traditional steno theories impose cognitive burdens that exceed what sustainable human performance can bear — massive brief memorization, phrase prediction, complex finger coordination under time pressure, all competing for the same limited mental resources simultaneously. Most of the people who never made it were not undisciplined or incapable. They were running a system that was never designed around how human brains actually function under load.

The attrition numbers have been visible for decades. The institutions that govern this profession had the data. The training systems were not redesigned at scale.

The reason is not difficult to identify: scarcity protected the survivors. A profession with a 90% dropout rate is, functionally, a profession in which high attrition keeps rates elevated for the 10% who make it through. Whether that dynamic was the product of conscious policy or structural inertia is a secondary question. The outcome is the same. The pipeline was never rebuilt. The workforce aged without replacement. And the shortage grew large enough and persistent enough that agencies had both the justification and the economic incentive to build a parallel supply chain around digital reporters and AI.

They did not wait for AI to mature. The self-inflicted shortage handed them the opening. AI arrived to a market the profession had already vacated.

Here is where the market actually goes.

The bottom tier — routine depositions, standard hearings, uncomplicated proceedings with clean audio — is already gone for freelance stenographers in many major markets. AI transcription on clean audio is now accurate enough, in many cases, to shift the economics of routine work. A scopist cleans what remains faster and at lower cost than a full stenographic transcript. The agency captures the margin that previously went to you. This is not a coming disruption. In many markets, it is already the default.

The middle tier — multi-party civil litigation, standard criminal proceedings, cases with moderate technical vocabulary — is vulnerable within five to ten years. Current AI struggles with overlapping speakers, heavy accents, and domain-specific terminology. That gap is closing every quarter. Agencies are building the infrastructure now — internal scoping teams, proprietary platforms, AI-assisted workflows — so they can scale the moment the economics cross the threshold. The preparation is not speculative. It is visible in the organizational structures being created right now.

The top tier — complex federal litigation, capital proceedings, high-stakes realtime, CART — survives. A human reporter can interrupt to clarify a mumbled answer, manage the dynamics of a difficult room, parse ambiguous context in real time, and produce a certified realtime feed that unreviewed AI output cannot legally substitute for under current federal and state rules. That capability remains irreplaceable in the most demanding proceedings. That tier will still exist in 20 years. It will be smaller, more specialized, and occupied by practitioners who function less like volume transcriptionists and more like high-stakes consultants.

If your income depends on routine deposition volume routed through large agencies, you are already in the contraction. The question is whether you know it yet.

The hardest truth is the one the institutions never stated plainly.

The dropout rate was not inevitable. A 90% failure rate is not a feature of the work itself — it is the predictable consequence of training methods that were never designed around the humans using them. Methods that measured speed at the expense of cognitive sustainability. Methods that optimized for a metric that told nothing about execution difficulty. Methods that produced a workforce too small and too fragile to withstand the arrival of a cheaper alternative.

Every student who dropped out was a person with professional aspirations, family expectations, time and money spent, now carrying debt and searching for a different direction. Not a statistic. A life.

There were enough of those lives — enough closed programs, enough reporters aging out without replacements — that the agencies had no rational choice but to build infrastructure around the shortage. They are not villains. They are businesses responding to a vacuum the profession created.

But the vacuum was not inevitable either.

The displacement was not imposed on you from outside. It was made possible from within — by training systems that were never redesigned, by institutions that had structural reasons to leave the pipeline broken, by a profession that measured the wrong thing for a century and called the casualties a dropout rate.

The reporters who make it through the next decade will be the ones writing at the level where machines still cannot go. Getting there means treating this profession as a craft that demands mastery — not a volume game that rewards survival.

The system will keep only what it cannot replace.

Tom Fernicola has been a court reporter for 37 years, specializing in complex litigation in New York City. He is the creator of BREVITY, a stenographic system designed around cognitive load science.