Medical transcription is one of the jobs most directly touched by AI. Speech recognition turns a doctor's dictation into text, and newer tools draft the whole visit note while the doctor talks with the patient. Transcriptionists increasingly edit those drafts rather than typing from scratch.
The short answer: AI is shrinking this job, not ending it yet. The ILO's 2025 index puts medical secretaries, the group that includes transcriptionists, in gradient 3, with a score of 0.53. The US Bureau of Labor Statistics projects a 4 percent decline for medical transcriptionists from 2025 to 2035, but still about 6,200 openings a year. Projections are estimates, not guarantees.
How exposed medical transcriptionists are
The International Labour Organization scores occupations' tasks from 0 (cannot be automated with generative AI) to 1 (could be done without a human operator), and groups them into gradients from 1 to 4, with 4 the most exposed; below them sit minimal exposure and not exposed (ILO Working Paper 140, 2025).
| Occupation (ISCO-08) | ILO category | Mean task score (0 to 1) |
|---|---|---|
| Medical Secretaries | Gradient 3 | 0.53 |
| Typists and Word Processing Operators | Gradient 4 | 0.65 |
The ILO scores groups of occupations, so these are the closest groups rather than a score for transcriptionists alone.
What the BLS projects
| Occupation | Projected change 2025 to 2035 | Openings a year |
|---|---|---|
| Medical transcriptionists | Decline 4% | About 6,200 |
The BLS says advances in speech recognition and natural language processing let physicians document visits in real time, reducing the need for transcriptionists, and make each transcriptionist more productive. All openings come from replacing workers who leave.
What AI is taking on
- First-draft transcripts from dictation.
- Visit notes drafted while the doctor talks with the patient.
- Formatting reports into the health record.
- Expanding common medical abbreviations.
What stays with the transcriptionist
- Editing: reviewing speech recognition drafts for accuracy, which the BLS lists as part of the job.
- Catching errors: spotting inconsistencies, wrong drug names or doses and missing details.
- Hard audio: accents, noise and specialist terms that software gets wrong.
- Checking AI help: drafted notes can be confidently wrong, which NIST calls confabulation (NIST AI 600-1). See checking AI output.
How the job is changing
The role is moving from typing to quality checking. People in it often move toward editing, health records, coding or scribe work. This page is about jobs and is not medical advice.
See also will AI replace medical scribes?, will AI replace medical billing and coding? and will AI replace court reporters?
Pay and how to get in
The US Bureau of Labor Statistics puts the median annual pay for medical transcriptionists at $40,410 in May 2025, against $50,980 for all workers. The typical education needed to enter is: postsecondary nondegree award. See how this job compares in AI exposure by pay and AI exposure by education level.
Questions and answers
Will AI replace medical transcriptionists?
It is shrinking the job. The ILO puts the closest group in gradient 3, and the BLS projects a 4 percent decline from 2025 to 2035.
Why are transcription jobs declining?
The BLS cites speech recognition and natural language processing that let physicians document visits in real time.
Are there still transcription openings?
Yes. The BLS projects about 6,200 openings a year, all from replacing workers who leave.
What transcription work stays human?
Editing AI drafts, catching errors, handling difficult audio and checking AI notes.