■ MODERATE RISK ■ Healthcare
AI will draft the paperwork and gamify the home exercises, but occupational therapy's essence — figuring out how this particular human gets dressed, cooks, and works again in their particular life — is contextual problem-solving with hands on. The profession bends toward hybrid care; it doesn't break.
“AI adapts therapy plans in real-time. Your encouragement still matters. Probably.”
Our AI replacement risk score — how we score jobs
Occupational therapists rebuild daily life after it's been interrupted. A caseload might span a stroke survivor relearning to button a shirt, a child with sensory processing issues who can't sit through class, a warehouse worker post-injury needing a graded return to lifting, and an elder whose home is an obstacle course of fall hazards. The work: functional assessments (often in homes, schools, and workplaces), designing activity-based interventions, fabricating custom hand splints, recommending and fitting adaptive equipment, training families and caregivers, and documenting all of it against insurer scrutiny that treats every visit as guilty until proven medically necessary.
AI's entry points are real but peripheral. Ambient scribes are dissolving the documentation burden — no small thing in a profession that spends huge fractions of the day on notes. Sensor- and camera-based tools track home-exercise adherence and motion quality; VR and gamified rehab keep patients engaged between sessions; predictive models help flag fall risk; and generative tools can draft treatment-plan boilerplate and home-program handouts. Telehealth OT expanded permanently after the pandemic for coaching-style interventions. All of this compresses the routine, repeatable slice of the job — exercise supervision and administrative labor — which is what our 38 is pricing in.
The center holds because occupational therapy is contextual by definition. Assessing why a client can't cook safely requires standing in their actual kitchen, watching compensation patterns, and negotiating with their pride; splinting requires hands; pediatric sensory work runs on rapport a screen degrades badly. Judgments blend biomechanics, cognition, psychology, and the client's environment — a fusion AI handles poorly because half the inputs never appear in any dataset. Demographics also argue for the profession: aging populations, stroke survival, and workplace-injury economics keep expanding demand while training pipelines lag. The realistic future is OTs supervising tech-augmented home programs across larger caseloads — more orchestration, less rep-counting, and the in-person visit reserved for what actually needs hands and eyes.
Automatability: our editorial assessment of current and near-term AI capability
A slow-transformation profession: documentation automation and remote monitoring are arriving now and mostly subtract drudgery; through the 2030s expect larger caseloads managed hybrid-style, with in-person visits concentrated on assessment and hands-on work. Elimination risk stays low — demand growth from aging populations outruns efficiency gains, and the contextual core resists automation past 2040.
Less than most healthcare roles. The job's core is contextual problem-solving in real environments — homes, kitchens, workplaces — plus hands-on skills like splinting, none of which AI reaches. What's automating is documentation, scheduling, and exercise monitoring, which mostly returns time. Demographic demand growth further cushions the profession.
Expect notes that write themselves from ambient recording, home programs monitored by apps and sensors between visits, and predictive flags for fall risk or non-adherence. That shifts the week toward assessments, hands-on interventions, and complex cases, with routine check-ins going remote. Caseloads likely grow as the administrative overhead shrinks.
The fundamentals are favorable: aging populations, strong projected demand, and a skill set — environmental assessment, adaptive problem-solving, therapeutic rapport — that sits in automation's blind spot. Come out of school fluent in telehealth and monitoring tech, because employers are building hybrid models and will prefer graduates who can run them.
They can replace some supervised repetition for motivated, straightforward patients, and payers will push that where it works. They can't assess a cluttered home's fall hazards, fabricate a splint, win over a sensory-avoidant child, or notice the cognitive change a family missed. Apps are becoming the homework; the therapist remains the diagnosis and the plan.