Why this work stays in the gym
Will AI replace adapted physical education specialists? The honest answer starts with where the job happens. Most of it happens in a gym, a pool, or on a playground, next to a student who needs a hand to balance, a cue repeated three different ways, or a game rebuilt on the spot so they can join in. Judging a child’s gross motor skills in person, then modifying equipment and drills to match what you saw, is a physical and social task. Software can suggest a drill. It cannot steady a student who is about to fall.
The second half of the job is paperwork and planning. Specialists write goals for individualized education programs, track progress against those goals, draft lesson plans, and explain all of it to parents, classroom teachers and therapists. That is where text tools already help, and where the job is quietly changing. Drafting takes less time than it used to. Meetings do not.
Scale matters here too. Federal data puts about 33,930 people in this occupation, with median pay of $76,580 and projected employment growth of 1.9% between 2025 and 2035 (BLS, 2025). That is a small, specialized field tied to student head counts and legal obligations, not to software budgets. When districts cut, they tend to stretch one specialist across more schools rather than hand the role to a machine. For how the headline figure above is built, see how the scoring works.
What AI handles, what it assists, and what stays with people
The share of task time that AI can take on its own is small here: 0%. The tasks that sit closest to that line are the written and administrative ones, such as turning assessment notes into draft IEP goal language and producing progress summaries for a file. These outputs still need a specialist to check them against what the student actually did in class.
Assisted work is the bigger story: 35% of task time. Lesson planning is a good example. A specialist can ask a model for activity variations for a student using a walker, then pick what is safe for that gym and that group. Video tools can flag movement patterns in a recorded session, which gives the specialist a second look at a student’s gait or throwing form. The judgment, and the liability, stay with the person. The coverage method explains how task time gets sorted this way.
Work that needs a person accounts for 65% of task time. That includes hands-on instruction and physical support during activity, and managing behavior and safety when a student is upset or overstimulated. It also covers the IEP meeting itself, where a specialist has to read a parent’s worry and negotiate a goal everyone will sign.
What the evidence actually shows
There is no direct test of AI against adapted physical education specialists doing this job. The evidence grade for quality parity is D, which is our label for not measured. That matters: without a measured comparison, no parity number belongs on this page, and anyone quoting one is guessing.
What would settle it is narrow and testable. First, a study comparing AI-drafted IEP motor goals with specialist-written goals, scored blind by qualified reviewers for accuracy and legal sufficiency. Second, a trial of automated movement assessment against in-person assessment for the same students. Third, measured outcomes when a tool handles session planning for a term. Until work like that exists, the quality parity method holds the number back rather than inventing one. You can see how AI chat tools answer this question in our what the AIs say list.
When the picture could change
Most likely between 2036 and 2058 (8 in 10 of our scenarios). The replacement year method explains what that window is based on.
Two things could pull the date earlier. One is better automated movement capture: cheap cameras and wearables that produce assessment-grade data a district trusts for reporting. The other is administrative consolidation, where planning and documentation tools cut enough desk time that one specialist covers more schools, which thins entry-level openings before it touches senior roles.
Two things push back the other way. A meaningful slice of this role is physical, and the robotics tier attached to it is dexterous humanoid work, which is the hardest and slowest class of machine to field in a crowded gym full of children. The other brake is legal and human: IEP law puts a named, qualified person behind every goal and every meeting, and parents expect that person in the room.
How to stay needed
Lean into the parts of the job that nothing else can hold. Direct, hands-on instruction and physical support during activity. Behavior and safety management when a session goes sideways. And the IEP meeting, where you translate motor data into something a parent and a classroom teacher can act on.
Two skills are worth building. First, assessment writing that stands up to scrutiny, so you can edit a machine draft fast instead of starting over. Second, practical knowledge of assistive technology and adaptive equipment, including what the tools get wrong for students with complex needs.
What to do: keep a running record of the student outcomes you changed, because that is the evidence a district weighs when it decides how many specialists it needs.
If you are weighing nearby roles, look at Special Education Teachers, Elementary School, Teaching Assistants, Special Education and Exercise Trainers and Group Fitness Instructors. You can put any two of them side by side on our job comparison tool, read the wider picture for teaching occupations and the education sector, or scan the jobs that most need a person.