Why the work stays close to people
Teaching communication is live work. An instructor stands in a room while a student delivers a speech, watches the shaking hands and the lost place in the notes, and decides in that second what to say. Coaching a nervous speaker and running a seminar discussion both depend on being present. You cannot grade a room’s attention from a transcript.
Plenty of the surrounding work is text, though, and text is where current models are strongest. Lecture outlines, reading lists, rubrics, discussion prompts, quiz banks and first-draft comments on written assignments can all be produced in seconds. That is the honest pressure here: not the classroom disappearing, but parts of the prep and marking load shrinking, which usually shows up first in how many graduate assistants and adjuncts a department hires.
Pay and headcount give some context. The Bureau of Labor Statistics puts US employment for this occupation at about 29,420 with median pay of $78,580, and projects growth of roughly 2.1% from 2025 to 2035 (BLS). That is a small, slow-moving field where budget decisions matter more than technology on its own. Our scoring method treats that task mix, not the job title, as the thing being measured.
What AI drafts, what it assists, and what it leaves alone
Start with the work AI can take end to end. 14% of the task time falls in that group: routine material generation and routine marking, such as building a quiz bank from a reading or scoring a short written response against a fixed rubric. Our coverage figure, which estimates how much of the task time AI can handle today, is 39 out of 100; the coverage method explains how that is built.
Then there is the assisted middle. 42% of the time sits where a model speeds a person up without finishing the job. Preparing lecture materials and giving written feedback on drafts both belong here. The model produces a usable draft; the instructor decides what is right for this cohort, this assignment and this student, and carries the responsibility for the grade.
Finally, the part that keeps the job a job. 44% of the task time needs a person in the room or on the record: live coaching of delivery, advising students on course and career choices, and the departmental work of curriculum design, committees and accreditation paperwork. Almost none of this needs a robot body, which is why the robotics requirement for this occupation is scored as none needed.
What has actually been tested
Not much, in this job specifically. Our evidence grade is D, which means there is no direct, published head-to-head test of an AI system against qualified communications instructors. Because the grade is D, we publish no parity number at all. A number without a test behind it would just be a guess dressed up.
What would settle it is narrow and doable: a blinded study where instructors and a model each grade the same student speeches and written assignments, scored for agreement with expert raters and, better still, for whether students who got AI feedback improved as much over a semester. Benchmarks on essay scoring are not the same thing as coaching a live presentation. The quality parity method sets out what we accept as a real comparison.
When the picture could shift
Most likely between 2034 and 2045 (8 in 10 of our scenarios). That window comes from our model, not from a prediction about any one department; the replacement year method explains what it measures.
Two things could pull it earlier. The first is cost: running a model over prep and first-pass feedback is cheap next to the annual cost of staffing a section, and the cost panel on this page shows how wide that gap is. The second is scale, because large online sections are where automated feedback gets tried first and where entry-level teaching jobs are thinnest.
Two things hold it back. Accreditation and course design still assume a named instructor who can defend a grade, and live oral assessment is awkward to automate without video judgment nobody has validated here. Institutional adoption is also slow; colleges change syllabi and assessment rules on a yearly cycle, not a quarterly one.
What to do: keep a written record of the work that is clearly yours, especially live coaching, advising and curriculum design.
How to stay needed in communications teaching
Lean into the three task areas that sit with people here. First, live delivery coaching, where you respond to a specific speaker in real time. Second, student advising, which is judgment about a person’s path, not information retrieval. Third, curriculum and assessment design, including the harder question of how to assess speaking and writing when students have models of their own.
Two skills travel well. One is assessment design that still works with AI in the room: oral defenses, in-class work, iterative drafts with visible revision history. The other is practical fluency with the tools, so you can set rules for a class and explain them without hand-waving. The career planning guide covers how that reads to an employer.
If you are weighing options, nearby teaching roles share much of this task mix: English language and literature teachers, art, drama and music teachers and business teachers. You can put any two of them side by side on the job comparison tool, see the wider group on the postsecondary teachers family page, check how the field looks across the education sector, or browse the jobs that most need a person list.