Occupational Therapy Aides

31-2012.00
Median wage $39,160/yr4,310 employed (US)Rank #686 of 923 scored · top 74% by substitution

Under close supervision of an occupational therapist or occupational therapy assistant, perform only delegated, selected, or routine tasks in specific situations. These duties include preparing patient and treatment room.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution20
Exposure17
Augmentation39

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

15 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

7%

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%20

panel mean rating 1.8/5 → substitution pressure 20/100

Technical feasibility todayw 20%14

panel mean rating 1.5/5 → substitution pressure 14/100

Cost vs. human wagew 15%16

panel mean rating 1.6/5 → substitution pressure 16/100

Adoption barriersw 20%inverted — strong barriers lower the score30

panel mean rating 3.8/5 (barrier strength) → substitution pressure 30/100

Sector adoption velocityw 10%14

panel mean rating 1.6/5 → substitution pressure 14/100

Task breakdown (15 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Perform clerical, administrative, and secretarial duties, such as answering phones, restocking and ordering supplies, filling out paperwork, and scheduling appointments.

76

CI 72–79 · exposure 75 · augmentation 88 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Healthcare organizations, including therapy clinics, are rapidly adopting scheduling software, document management systems, and basic chatbots for administrative tasks. These are information-work sectors with strong digitization trends and proven ROI.
Sector adoption velocityclaude-sonnet-53/5Healthcare administrative functions are adopting AI scheduling/chat tools at a moderate pace, behind pure information/finance sectors but ahead of hands-on clinical care due to lower stakes.
Augmentation potentialclaude-haiku-4-5-202510014/5AI significantly assists aides in these tasks: scheduling software reduces manual calendar juggling, document processing speeds form completion, and phone systems filter calls and provide talking points. The aide remains in control while productivity gains are substantial.
Augmentation potentialclaude-sonnet-55/5AI scheduling assistants, automated reminders, and voice-based intake systems substantially reduce time aides spend on clerical tasks, freeing them for more patient-facing duties.
Task automatabilityclaude-haiku-4-5-202510014/5Most clerical and administrative duties can be substantially automated: phone answering via IVR/chatbots, supply ordering via inventory systems, form filling via document processing and data entry automation, and appointment scheduling via calendar integration and booking systems. However, some context-dependent judgment (e.g., prioritizing urgent calls, handling exceptions) may still require human oversight, preventing a full 5.
Task automatabilityclaude-sonnet-54/5Scheduling, phone answering, paperwork, and supply ordering are largely routine administrative tasks that current AI tools (chatbots, scheduling assistants, RPA) can handle end-to-end for a majority of instances.
Adoption barriersclaude-haiku-4-5-202510012/5Few legal or regulatory barriers exist for automating clerical work; occupational therapy practice itself requires human judgment, but administrative support does not. Main barriers are organizational inertia and customer/staff preference for human contact on some interactions, but these are not hard constraints.
Adoption barriersclaude-sonnet-52/5No licensing requirement for clerical work, but healthcare settings often require HIPAA-compliant systems and some patient-facing tasks benefit from human judgment for scheduling nuances or urgent issues.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-based solutions for these routine administrative tasks (cloud-based scheduling, RPA for forms, chatbots for phone screening) cost a small fraction of a full-time aide's wage; off-the-shelf products easily achieve order-of-magnitude cost savings per task unit.
Cost vs. human wageclaude-sonnet-54/5Automated scheduling and phone/chat systems cost a fraction of aide wages per interaction, though integration with clinic EHR/practice management systems adds some overhead.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed products reliably handle individual components: phone systems with AI screening, inventory management platforms, document processing (OCR, form parsing), and automated scheduling tools. These are widely used in healthcare settings, though integration across all duties and handling edge cases may require customization and oversight.
Technical feasibility todayclaude-sonnet-54/5AI scheduling assistants, virtual receptionists, and inventory management systems are deployed at scale across healthcare and other industries today, though clinic-specific integration and edge cases still require human oversight.

Prepare and maintain work area, materials, and equipment and maintain inventory of treatment and educational supplies.

33

CI 30–35 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Occupational therapy is a healthcare sector with moderate digital maturity; inventory systems see some adoption, but physical preparation automation remains minimal and adoption is slow in this fragmented, often small-practice setting.
Sector adoption velocityclaude-sonnet-52/5Healthcare support occupations are slow adopters of AI for physical/logistical tasks, though inventory software adoption in clinics is moderate.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-driven inventory forecasting and tracking systems can assist aides in prioritizing supply replenishment and organizing materials more efficiently. Augmentation is moderate—it streamlines logistics but does not transform the core manual labor.
Augmentation potentialclaude-sonnet-53/5AI-based inventory management and supply-ordering tools can meaningfully assist with tracking and reordering supplies, improving efficiency of that sub-task.
Task automatabilityclaude-haiku-4-5-202510012/5Physical preparation and maintenance of work areas requires in-situ manipulation and environmental assessment that current AI cannot perform autonomously. Inventory tracking is automatable, but the physical setup and upkeep components that dominate this task remain largely manual.
Task automatabilityclaude-sonnet-52/5Involves physical tasks like arranging equipment and materials in a clinical space, which current AI cannot perform; only the inventory-tracking/ordering portion is automatable.
Adoption barriersclaude-haiku-4-5-202510013/5Some organizational friction exists around adopting inventory software, but no hard legal barriers prevent automation of supply tracking. Physical workspace preparation has moderate friction due to the need for context-sensitive decision-making in therapeutic environments.
Adoption barriersclaude-sonnet-52/5No licensing requirement for this specific task, but it occurs within a clinical setting requiring physical presence and coordination with therapists.
Cost vs. human wageclaude-haiku-4-5-202510012/5Inventory software is inexpensive, but the per-unit cost of deploying robotics or autonomous systems for physical workspace preparation and material handling would exceed the wage of a part-time or entry-level aide in most settings.
Cost vs. human wageclaude-sonnet-52/5Physical setup still requires a human aide on-site; only digital inventory tracking could be cheaper, so blended cost is not favorable to AI.
Technical feasibility todayclaude-haiku-4-5-202510012/5Inventory management software exists and is deployed, but reliable autonomous physical preparation and maintenance of therapeutic spaces remains research-stage. Real-world occupational therapy settings still rely on human aides for these hands-on functions.
Technical feasibility todayclaude-sonnet-52/5Inventory management software exists and is deployed, but the physical preparation of work areas and equipment has no viable AI product today.

Report to supervisors or therapists, verbally or in writing, on patients' progress, attitudes, attendance, and accomplishments.

32

CI 25–39 · exposure 33 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare remains a laggard in AI adoption relative to information and finance sectors due to regulatory overhead, liability concerns, and organizational conservatism around patient data and clinical documentation.
Sector adoption velocityclaude-sonnet-52/5Healthcare support occupations, especially aide-level roles, show slower AI adoption compared to office/professional services due to physical presence requirements and regulatory environment.
Augmentation potentialclaude-haiku-4-5-202510013/5AI tools can assist aides by drafting summaries from session notes and flagging attendance or behavioral patterns, materially reducing time spent on documentation while the aide retains final responsibility for accuracy and clinical judgment.
Augmentation potentialclaude-sonnet-53/5AI can help structure, summarize, and draft written reports based on aide's observations, improving efficiency of the documentation portion of this task.
Task automatabilityclaude-haiku-4-5-202510013/5AI can assist with structured data entry and generate draft progress reports from patient session notes, but the task requires human judgment about patient attitudes and behavioral nuances that current systems handle inconsistently. A meaningful portion of reporting could be automated, but quality parity typically requires human review and revision.
Task automatabilityclaude-sonnet-52/5Drafting written progress reports from structured notes could be AI-assisted, but the underlying observation of patient attitudes/behavior and verbal reporting requires human presence and judgment, so end-to-end automation is limited.
Adoption barriersclaude-haiku-4-5-202510014/5Healthcare liability, patient privacy (HIPAA), and clinical documentation standards create substantial friction; supervisors and therapists typically require human accountability for progress documentation, and insurance/accreditation bodies often mandate human sign-off on clinical records.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement for this specific reporting task, but clinical documentation often has oversight and accountability norms tied to patient care records.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integration of AI tools into clinical workflows adds licensing, oversight, and liability costs that partially offset wage savings; the human aide's loading and AI costs remain roughly comparable given compliance requirements in healthcare.
Cost vs. human wageclaude-sonnet-52/5Human aides are relatively low-wage, and AI would need integration with observation and communication workflows, so cost savings are modest rather than dramatic.
Technical feasibility todayclaude-haiku-4-5-202510012/5While NLP systems can extract and summarize clinical notes, no mainstream product reliably generates compliant occupational therapy progress reports that supervisors would accept without substantial revision. Clinical documentation still primarily relies on human-authored records in practice.
Technical feasibility todayclaude-sonnet-52/5AI documentation tools exist in healthcare settings (e.g., clinical note summarization) but are not widely deployed specifically for aide-level verbal/written progress reporting to therapists.

Observe patients' attendance, progress, attitudes, and accomplishments and record and maintain information in client records.

28

CI 25–30 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of clinical AI remains cautious and pilot-heavy; occupational therapy settings, which are often smaller and distributed, lag behind high-digitization sectors in actual production deployment of autonomous documentation systems.
Sector adoption velocityclaude-sonnet-52/5Healthcare and allied health support occupations have historically been slower to adopt AI compared to information/finance sectors, with EHR-linked AI scribes still in early-to-moderate deployment.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist aides by auto-transcribing observations, suggesting documentation structure, and flagging missing data, thereby improving note quality and efficiency; however, the human aide remains essential for the actual observation and judgment.
Augmentation potentialclaude-sonnet-53/5AI-assisted documentation tools (e.g., ambient scribes, summarization) can meaningfully speed up record-keeping and note drafting for aides, though the core observational judgment remains human-driven.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with structured record-keeping and data entry from notes, but observing patients' attitudes, progress assessment, and accomplishments requires human judgment and contextual understanding that current systems cannot reliably perform end-to-end with 50%+ time savings at equal quality.
Task automatabilityclaude-sonnet-52/5Recording and maintaining structured information can be partially automated, but the core act of observing patient attitudes, progress, and accomplishments requires in-person human judgment and physical presence that current AI cannot replicate.'
Adoption barriersclaude-haiku-4-5-202510014/5Healthcare regulatory requirements (HIPAA, clinical documentation standards, licensing of therapeutic staff) and liability concerns create strong barriers; clinical records often require a licensed or supervised professional to validate and sign off on patient progress assessments.
Adoption barriersclaude-sonnet-53/5No licensing requirement strictly limits documentation to a therapist, but healthcare privacy regulations, accuracy requirements for medical records, and organizational policies create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI transcription and documentation tools are relatively inexpensive, but the need for human observation and clinical judgment means the human aide cannot be fully replaced; integration costs and oversight maintain rough parity with aide wages.
Cost vs. human wageclaude-sonnet-52/5AI note-taking/transcription tools are cheap relative to aide wages for the documentation portion, but since observation and clinical judgment still require the human on-site, overall cost savings are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5While basic documentation software exists, no deployed product reliably observes and autonomously interprets patient attitudes, progress, and accomplishments without human oversight; systems require substantial manual data entry and clinical validation.
Technical feasibility todayclaude-sonnet-52/5Documentation/transcription tools exist and are used in clinical settings, but no deployed product independently observes and assesses patient attitudes and accomplishments reliably in production.

Manage intradepartmental infection control and equipment security.

23

CI 20–25 · exposure 20 · augmentation 38 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare settings have been slow to adopt AI for operational management tasks beyond diagnostic imaging; infection control and security remain largely manual and human-centric despite digitization opportunities.
Sector adoption velocityclaude-sonnet-52/5Healthcare support occupations, especially allied health aide roles, show slow AI adoption for physical safety and compliance tasks compared to purely informational sectors.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could meaningfully assist by automating infection control checklists, flagging equipment overdue for maintenance, and analyzing security logs, but the human aide must retain primary responsibility for physical verification and corrective action.
Augmentation potentialclaude-sonnet-52/5AI can support with checklists, reminders, or IoT-based equipment tracking systems, offering modest assistance, but the core inspection, judgment, and enforcement remain manual.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist with infection control documentation and monitoring protocols, the task requires real-time physical oversight of equipment security, staff compliance, and environmental conditions that demand human presence and decision-making in clinical settings. Most of the core work cannot be meaningfully automated today.
Task automatabilityclaude-sonnet-52/5This task blends physical equipment checks, sanitation enforcement, and situational judgment about security that current AI cannot execute end-to-end; only administrative sub-components like tracking logs could be partially automated.rapport.
Adoption barriersclaude-haiku-4-5-202510014/5Healthcare infection control and equipment security are heavily regulated by HIPAA, OSHA, and accreditation bodies that often require documented human accountability and inspection; liability and patient safety considerations create hard barriers to full automation.
Adoption barriersclaude-sonnet-54/5Infection control protocols are governed by healthcare regulations and require accountable human compliance, and equipment security often involves liability and physical custody responsibilities that resist full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of deploying AI systems for comprehensive monitoring (sensors, cameras, analytics platforms, integration with security systems) plus required human oversight would likely exceed the loaded wage of an aide performing these duties today.
Cost vs. human wageclaude-sonnet-52/5Any AI contribution would be limited to a monitoring or logging layer alongside human oversight, so overall cost is likely comparable to or higher than the human aide performing the full physical task.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed products reliably perform integrated infection control and equipment security management end-to-end in healthcare settings; existing systems handle narrow components (e.g., checklist reminders) but lack the contextual judgment and physical verification requirements of this task.
Technical feasibility todayclaude-sonnet-51/5There are no deployed AI products performing intradepartmental infection control or physical equipment security management in clinical/therapy settings today; this remains a human physical-facility management task.

Instruct patients and families in work, social, and living skills, the care and use of adaptive equipment, and other skills to facilitate home and work adjustment to disability.

15

CI 5–25 · exposure 13 · augmentation 63 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption is slow: occupational therapy is a small, highly regulated clinical sector with strong preference for human interaction and minimal financial incentive for rapid automation. Most settings remain small clinics or institutional providers where organizational friction and licensing requirements limit AI deployment.
Sector adoption velocityclaude-sonnet-52/5Healthcare/rehab settings are typically slower AI adopters for hands-on care tasks, with pilots limited mostly to documentation and administrative support rather than direct patient instruction.
Augmentation potentialclaude-haiku-4-5-202510014/5AI could significantly augment aides by generating customized instructional materials, demonstrating adaptive equipment use via video, tracking patient progress, and organizing learning plans—allowing the human aide to focus on live engagement, motivation, and real-time problem-solving with patients and families.
Augmentation potentialclaude-sonnet-53/5AI can help generate educational materials, checklists, and personalized instructional content or reminders that aides use to supplement in-person teaching, offering moderate productivity support.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could generate instructional content and adaptive equipment guidance, the task fundamentally requires real-time, personalized interaction with patients and families to assess comprehension, adjust teaching methods, and build the therapeutic relationship necessary for behavioral and skill change. Current AI systems cannot reliably conduct this live, adaptive, emotionally responsive instruction at scale.
Task automatabilityclaude-sonnet-51/5This requires in-person, hands-on demonstration, physical assistance, and adaptive real-time coaching tailored to a patient's specific physical/cognitive limitations, which current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Occupational therapy aides work under the supervision and authorization of licensed occupational therapists; liability for patient outcomes, adaptive equipment safety, and skill transfer rests with the clinical team. Regulatory frameworks (state licensure, accreditation standards) require qualified human oversight of patient instruction and care.
Adoption barriersclaude-sonnet-54/5Occupational therapy aides work under supervision of licensed therapists with liability, safety, and often regulatory/insurance requirements tied to hands-on patient instruction, creating substantial barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Generating instructional content is cheap, but oversight, customization, and the need for human clinical staff to validate, personalize, and deliver instruction to patients keeps total cost per patient outcome comparable to or higher than direct aide labor, especially in supervised settings.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical instruction and supervision involved, so there is no viable AI cost comparison for full task replacement.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI can produce instructional materials and video demonstrations of adaptive equipment use, but no deployed product reliably performs the full task of assessing individual patient needs, delivering personalized skill instruction, and monitoring progress in real clinical settings. Chatbots and video content lack the situational judgment and real-time responsiveness required.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently instructs patients on adaptive equipment use or living skills in a clinical/home setting; this remains a human-delivered service.

Demonstrate therapy techniques, such as manual and creative arts and games.

15

CI 5–25 · exposure 13 · augmentation 38 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare and therapy settings have adopted AI for administrative tasks and some assessment, but hands-on therapeutic technique demonstration remains largely traditional and human-delivered. Adoption of automation in direct patient-facing, physical therapy contexts has been very slow.
Sector adoption velocityclaude-sonnet-51/5Healthcare aide roles involving hands-on physical care are among the slowest sectors for AI adoption due to physical, regulatory, and safety constraints.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could usefully assist by providing video references, suggested therapeutic sequences, or creative-activity templates that an aide reviews before or during live demonstrations, thereby improving confidence and consistency without replacing the human's embodied presence and real-time adjustment.
Augmentation potentialclaude-sonnet-52/5AI could help plan or suggest therapy activities and track progress, but offers minimal assistance to the actual physical demonstration and hands-on interaction central to this task.
Task automatabilityclaude-haiku-4-5-202510012/5Demonstrating therapy techniques requires real-time, embodied interaction with patients—manual handling, observation of response, and adaptive adjustment. While AI could produce instructional videos or verbal guidance, hands-on manual demonstration and live creative engagement cannot be meaningfully automated by current systems, and no AI system can substitute for the physical presence and responsiveness central to effective therapy techniques.
Task automatabilityclaude-sonnet-51/5This requires physical, in-person demonstration of manual therapy techniques and hands-on guidance with patients, which current AI cannot perform end-to-end given the physical embodiment required.
Adoption barriersclaude-haiku-4-5-202510014/5Occupational therapy is a regulated healthcare field; physical handling and patient contact typically fall under scope-of-practice rules requiring a licensed OT or supervised aide. Liability exposure for incorrect or unsafe demonstration, combined with patient-contact requirements, creates substantial regulatory and legal barriers to substitution.
Adoption barriersclaude-sonnet-54/5This involves direct patient care requiring physical presence, safety supervision, and often occurs under licensed therapist oversight, creating strong practical and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of deploying AI video, agents, or simulation systems for therapy demonstration, plus human oversight to ensure patient safety and clinical appropriateness, approaches or exceeds the cost of a direct human aide demonstrating techniques in real time.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for the physical demonstration component, so the human remains the only cost-effective option for this task.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI systems today cannot reliably perform live therapeutic demonstrations involving manual techniques, physical touch, or nuanced real-time adaptation to patient response. Video generation and chatbot instruction exist but do not meet production standards for actual therapy delivery; a licensed or supervised human remains operationally required.
Technical feasibility todayclaude-sonnet-51/5No deployed product can physically demonstrate manual therapy techniques or supervise patients through creative arts/games activities in a clinical setting today.

Assist educational specialists or clinical psychologists in administering situational or diagnostic tests to measure client's abilities or progress.

15

CI 5–25 · exposure 13 · augmentation 38 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare and educational psychology remain low-digitization, conservative sectors with strong preference for human contact in assessments; pilot automation is rare and adoption of autonomous test administration is minimal.
Sector adoption velocityclaude-sonnet-51/5Healthcare and educational support aide roles involving physical client interaction show minimal AI agent adoption in production settings.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by pre-populating demographic data, auto-scoring objective items, and flagging outlier responses for clinician review, raising aide productivity modestly, though the core interpersonal and observational task remains human-centered.
Augmentation potentialclaude-sonnet-52/5AI could help with scheduling, scoring standardized tests, or generating reports post-assessment, but offers little assistance during the hands-on administration process itself.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could handle standardized test administration scripts and basic scoring of objective items, the task requires careful behavioral observation, judgment about client comfort/engagement, and real-time adaptation to client responses—elements current systems cannot reliably perform end-to-end without substantial human oversight, preventing the 50% time-saving threshold.
Task automatabilityclaude-sonnet-51/5This requires physical presence, hands-on assistance with clients (often with disabilities), and real-time adaptive support during testing that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Administering diagnostic tests is legally and ethically governed in clinical and educational settings; psychological test administration often requires credentialing oversight, client confidentiality compliance, and liability exposure if errors occur—all imposing material friction on unsupervised automation.
Adoption barriersclaude-sonnet-54/5Clinical testing often requires supervision by licensed psychologists or specialists, involves vulnerable populations, and has liability/regulatory considerations that favor human involvement.
Cost vs. human wageclaude-haiku-4-5-202510012/5The human aide wage is modest (typically $25K–$35K annually), and any AI system capable of actual test administration would require significant setup, training data, and oversight infrastructure that would exceed the cost advantage over a junior aide.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical and interpersonal assistance role, so there is no viable cost comparison—human labor is required.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed products reliably administer diagnostic psychological or educational tests independently; existing AI can score some standardized assessments from data, but administering tests requires human presence, rapport-building, and judgment that remains largely research-stage.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product assists physically or interpersonally in administering situational/diagnostic tests in clinical or educational settings; this remains a human aide role.

Adjust and repair assistive devices and make adaptive changes to other equipment and to environments.

14

CI 10–19 · exposure 8 · augmentation 25 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Occupational therapy remains a labor-intensive, client-centered field with limited digitization; adoption of AI or robotics for physical adaptation work is negligible in production settings today.
Sector adoption velocityclaude-sonnet-51/5Healthcare support occupations involving physical equipment adjustment are among the slowest to adopt AI, given the physical and manual nature of the work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with documentation, generating adaptation ideas, or providing educational resources about device modifications, but the core physical and assessment work remains human-dependent; limited opportunity for meaningful productivity gains.
Augmentation potentialclaude-sonnet-52/5AI could potentially assist with diagnostics, documentation, or suggesting adaptive modifications, but it offers minimal help with the actual physical adjustment and repair work.
Task automatabilityclaude-haiku-4-5-202510012/5This task requires physical manipulation, spatial assessment, and individualized problem-solving that current AI cannot perform end-to-end. While AI could potentially assist with documentation or identifying adjustment needs via images, the actual repair and environmental adaptation demands hands-on work, tool use, and real-world testing that only humans can accomplish today.
Task automatabilityclaude-sonnet-51/5This is a hands-on physical task involving manual adjustment, repair, and fitting of devices in physical spaces, which current AI systems cannot perform end-to-end.atan
Adoption barriersclaude-haiku-4-5-202510013/5While occupational therapy aides are not licensed professionals, the task involves safety-critical modifications to medical equipment and personal environments that require human judgment, accountability, and direct client interaction; some regulatory and liability concerns exist around equipment modifications.
Adoption barriersclaude-sonnet-53/5While not formally licensed in most jurisdictions, the task requires physical presence, hands-on skill, and judgment about patient safety and equipment fit, creating strong practical barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems capable of physical manipulation and adaptation work do not exist in deployable form, so the comparison is not meaningful; where physical robots might theoretically help, they remain vastly more expensive than trained human aides.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical labor and manual dexterity required, so there is no viable AI cost comparison for this task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI products can physically adjust, repair, or adapt equipment and environments without direct human operation. Although robotics research exists, no production systems currently perform this work reliably in occupational therapy settings.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs physical repair or environmental modification of assistive equipment; this remains firmly in the domain of human technicians.

Evaluate the living skills and capacities of clients with physical, developmental, or mental health disabilities.

11

CI 5–18 · exposure 8 · augmentation 25 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare and rehabilitation settings have been slow to adopt AI automation in direct client assessment. The sector is characterized by small organizations, strong human-contact requirements, and regulatory conservatism around clinical evaluation, resulting in minimal AI adoption in functional capacity assessment.
Sector adoption velocityclaude-sonnet-51/5Healthcare aide roles involving physical, hands-on assessment are a low-digitization, slow-adopting sector for AI-driven task automation.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could offer modest assistance in documenting standardized scores or organizing assessment data post-observation, but it provides minimal augmentation during the actual evaluation process, which depends on the aide's direct observation, interpersonal skills, and ability to interpret adaptive responses in context.
Augmentation potentialclaude-sonnet-52/5AI could help with documentation, scheduling, or summarizing observations after the fact, but offers minimal assistance to the actual hands-on evaluation process itself.
Task automatabilityclaude-haiku-4-5-202510012/5Evaluating living skills and capacities requires nuanced observation of functional abilities, adaptive responses, and individualized judgment that depends heavily on direct client interaction. While AI could assist in standardized questionnaire scoring or data organization, it cannot reliably assess the complex, context-dependent adaptive behaviors and physical demonstrations that occupational therapy aides must observe in real time.
Task automatabilityclaude-sonnet-51/5Evaluating living skills requires hands-on observation, physical interaction, and clinical judgment about a client's disability status, which current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: occupational therapy aides operate under state licensure frameworks and organizational protocols requiring human accountability for client evaluations. Clinical liability for incorrect functional assessments is severe, and many jurisdictions require licensed professionals to validate or directly conduct functional evaluations, creating regulatory and institutional friction.
Adoption barriersclaude-sonnet-54/5Client evaluation typically falls under supervision of licensed occupational therapists with documentation and liability requirements, creating strong professional and regulatory barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of AI systems capable of observing and evaluating functional capacity (if they existed) would be high relative to the wage of an aide, especially since human oversight of any assessment output would still be required for clinical validity.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this evaluation, so any AI cost would be additive to, not replacing, the human labor cost.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs end-to-end functional capacity evaluation of clients with disabilities. This task requires in-person observation, rapport-building, and real-time clinical judgment—capabilities that remain at research stage and have not achieved production-scale reliability in occupational therapy settings.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently conducts client evaluations of living skills; this remains a human clinical task performed in-person by trained aides/therapists.

Encourage patients and attend to their physical needs to facilitate the attainment of therapeutic goals.

9

CI 5–14 · exposure 8 · augmentation 25 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a conservative, heavily regulated sector with low automation of direct patient-facing roles. Adoption of AI in occupational therapy settings is minimal; human aides remain the standard of care.
Sector adoption velocityclaude-sonnet-51/5Healthcare aide roles involving physical patient contact are among the slowest to adopt AI/automation, given the physical, low-digitization nature of the work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist with scheduling, progress tracking, or reminding aides of therapeutic goals, but it offers limited augmentation to the core task of real-time encouragement and physical assistance, which is fundamentally interpersonal and kinesthetic.
Augmentation potentialclaude-sonnet-52/5AI could offer minor support like scheduling, tracking progress notes, or motivational content suggestions, but does not meaningfully augment the physical, interpersonal core of this task.
Task automatabilityclaude-haiku-4-5-202510012/5This task requires sustained, responsive human interaction and physical care—encouraging patients emotionally and assisting with bodily needs. Current AI cannot reliably perform these interpersonal and physical components end-to-end; a human must remain central to motivation and hands-on support.
Task automatabilityclaude-sonnet-51/5This task requires direct physical assistance, emotional encouragement, and hands-on care of patients that current AI cannot perform end-to-end.dressing, transferring, and comforting patients requires embodied physical presence.
Adoption barriersclaude-haiku-4-5-202510014/5Occupational therapy aides typically work under clinical/regulatory oversight, and direct patient care (physical assistance and emotional support) usually requires a licensed or credentialed human present. Liability and regulatory frameworks strongly protect this role from automation.
Adoption barriersclaude-sonnet-54/5Direct physical patient contact, safety concerns, and healthcare facility requirements for trained personnel create strong practical and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI tools that might offer communication reminders or scheduling aids are marginal supplements; they do not substitute for the aide's labor. The human wage remains far lower than the integrated cost of robotic systems and oversight needed.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing physical patient care, so cost comparison favors the human aide entirely; robotics for this remains experimental and expensive.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can autonomously encourage patients or attend to physical needs in a clinical setting. Robotic assistance exists for limited mobility tasks but cannot replace the relational and adaptive care this task demands.
Technical feasibility todayclaude-sonnet-51/5No deployed product physically attends to patients' needs or provides in-person motivational support during therapy sessions; this remains outside current AI product capability.

Assist occupational therapists in planning, implementing, and administering therapy programs to restore, reinforce, and enhance performance, using selected activities and special equipment.

8

CI 0–16 · exposure 8 · augmentation 25 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare occupational therapy is a traditionally low-digitization sector with strong human-contact requirements. Adoption of AI for aide-level therapeutic tasks remains minimal, with no evidence of AI agent deployment in production OT settings.
Sector adoption velocityclaude-sonnet-52/5Healthcare support occupations, especially those involving physical patient care, show slow AI adoption compared to office-based information work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could modestly assist with documentation, scheduling activities, or tracking patient progress data, but the therapeutic intervention itself—selecting, demonstrating, and adapting activities in real time with patients—is not meaningfully augmented by current AI tools.
Augmentation potentialclaude-sonnet-52/5AI can help with scheduling, documentation, and treatment plan suggestions, but offers limited assistance for the core physical and interpersonal components of the task.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist with some administrative elements (scheduling, documentation), the core of this task—implementing and administering therapy programs that require real-time physical interaction, individualized activity selection, and responsiveness to patient needs—demands human presence and judgment that cannot be meaningfully automated. Current systems cannot reliably perform the hands-on therapeutic activities or adaptive decision-making required.
Task automatabilityclaude-sonnet-51/5This requires hands-on physical assistance, direct patient handling, and adaptive real-time judgment during therapy sessions that current AI systems cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Occupational therapy aide roles are either directly regulated or work under the direct supervision of licensed occupational therapists. Patient safety, physical hands-on care, and the requirement for human professional oversight create hard barriers to substitution with automation.
Adoption barriersclaude-sonnet-54/5Healthcare settings require supervised, often credentialed personnel for patient safety, and liability concerns around physical therapy assistance create strong barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems cannot perform the core therapeutic and physical-assistance functions of an OT aide, so there is no cost-equivalent comparison. Where AI might assist (documentation, scheduling), the integration overhead and remaining human supervision would not achieve cost savings compared to an aide's loaded wage.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical labor and supervision involved, so there is no viable cost comparison—human labor remains the only functional option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs occupational therapy aide work end-to-end. The task requires physical presence with patients, real-time therapeutic judgment, and adaptive modification of activities based on patient feedback—beyond the scope of current AI systems in production healthcare settings.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical therapy assistance or equipment administration with patients; this remains firmly a human physical-presence task.

Supervise patients in choosing and completing work assignments or arts and crafts projects.

7

CI 0–14 · exposure 8 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare, especially occupational therapy settings, shows slow AI adoption for patient-facing roles. Regulatory conservatism, union presence in many facilities, and patient-contact requirements create structural resistance to automation in this domain.
Sector adoption velocityclaude-sonnet-51/5Healthcare support/aide roles involving hands-on patient supervision show minimal AI adoption in production settings.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by recommending appropriate projects, tracking patient progress notes, or flagging safety concerns, helping aides organize their work—but augmentation is modest because the core task (direct supervision and behavioral response) remains human-dependent.
Augmentation potentialclaude-sonnet-52/5AI could help plan or suggest activity options or track progress notes, but offers little assistance during the actual real-time supervision itself.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could help schedule or suggest projects, supervising patients meaningfully requires real-time observation, behavioral assessment, and adaptive response to emotional/physical needs—tasks that demand present human judgment and interpersonal sensitivity. Current systems cannot reliably handle the unpredictable, individualized nature of patient supervision at equal quality.
Task automatabilityclaude-sonnet-51/5This requires physical presence, real-time observation of patient behavior and abilities, and adaptive interpersonal supervision that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Healthcare regulations and liability frameworks require a qualified human to directly supervise patients during therapeutic activities. Legal and accreditation standards mandate human accountability for patient safety, behavior management, and therapeutic progress—creating hard barriers to autonomous AI substitution.
Adoption barriersclaude-sonnet-54/5Direct patient supervision in a therapeutic context typically requires trained staff present for safety, liability, and care-quality reasons, creating strong organizational and regulatory friction against automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5An occupational therapy aide's loaded cost is modest ($30–35k/year), and the task requires continuous on-site human presence; AI supervision systems would require significant infrastructure, oversight, and liability insurance—making total cost higher than human labor.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical supervisory task, so the human remains the only cost-effective option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably supervises patients in therapeutic or work settings. AI systems lack the embodied presence, safety oversight capability, and liability-acceptable autonomy to perform this task in production healthcare environments.
Technical feasibility todayclaude-sonnet-51/5No deployed product supervises patients in-person during therapeutic activities; this remains firmly outside current AI product capability.

Transport patients to and from the occupational therapy work area.

3

CI 0–5 · exposure 0 · augmentation 13 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a laggard sector for embodied automation; patient transport automation has shown minimal real-world adoption despite pilot projects, due to safety and regulatory constraints.
Sector adoption velocityclaude-sonnet-51/5Healthcare support/aide roles involving physical patient handling are among the least digitized and slowest to adopt AI or robotic automation.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist minimally—e.g., route optimization or scheduling logistics—but the core task of physically moving vulnerable patients remains human-centered with limited augmentation potential.
Augmentation potentialclaude-sonnet-51/5AI offers essentially no meaningful assistance for the physical act of transporting a patient; this is a manual, hands-on task outside current AI capabilities.
Task automatabilityclaude-haiku-4-5-202510011/5Transporting patients requires physical mobility, spatial navigation, and real-time responsiveness to patient needs and environmental hazards. Current AI cannot safely and reliably perform this embodied task end-to-end without dedicated robotic infrastructure that is not yet standard in clinical settings.
Task automatabilityclaude-sonnet-51/5Physically transporting patients (e.g., via wheelchair) requires embodied physical manipulation and safety judgment that current AI systems, including robots, cannot perform reliably or affordably in clinical settings today.
Adoption barriersclaude-haiku-4-5-202510015/5Patient safety, liability, regulatory oversight (HIPAA, medical devices), and the legal requirement for human attendants during patient transport create hard adoption barriers. Direct human contact and duty of care are fundamental to this task.
Adoption barriersclaude-sonnet-54/5Direct physical handling of patients involves safety, liability, and often regulatory/accreditation requirements for trained personnel, creating strong barriers against non-human automation even if technology existed.
Cost vs. human wageclaude-haiku-4-5-202510011/5Deploying autonomous transport robots or mobile manipulation systems for patient transport is vastly more expensive than the loaded wage of an occupational therapy aide, with high integration and liability costs.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI/robotic solution deployed at scale for this task, so the aide's wage remains far cheaper than any hypothetical automation requiring specialized hardware, safety certification, and maintenance.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed commercial system reliably transports human patients in healthcare environments today. While some hospitals pilot autonomous transport, these are narrow-scope pilots with extensive infrastructure setup and remain rare in production.
Technical feasibility todayclaude-sonnet-51/5No deployed commercial product autonomously transports patients within healthcare/rehab facilities; patient-moving robots remain research or pilot-stage with very limited scope.

Accompany patients on outings, providing transportation when necessary.

0

CI 0–0 · exposure 0 · augmentation 13 · importance 3.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare, particularly care for vulnerable populations, remains a sector with slow AI adoption due to regulatory caution, liability concerns, and high stakes of errors. No measurable displacement of human aides by AI systems is occurring in this task.
Sector adoption velocityclaude-sonnet-51/5Healthcare aide and patient transport roles are physical, low-digitization work with minimal AI adoption or displacement to date.
Augmentation potentialclaude-haiku-4-5-202510011/5AI cannot meaningfully augment this task because its core value is human presence, supervision, and real-time adaptive support in unpredictable situations. No AI tool substantially enhances the aide's ability to accompany or care for patients during outings.
Augmentation potentialclaude-sonnet-52/5AI could help with route planning, scheduling, or reminders around outings, but offers little assistance to the core physical accompaniment task itself.
Task automatabilityclaude-haiku-4-5-202510011/5This task involves physical presence, real-time responsiveness, and individualized human interaction in unpredictable environments (outings, transportation). Current AI systems cannot operate vehicles autonomously in most jurisdictions, nor can they provide the personalized behavioral support, assistance, or supervision that patients require during community outings.
Task automatabilityclaude-sonnet-51/5This is a physical accompaniment and transportation task requiring bodily presence, driving, and real-time responsiveness to a patient's physical/medical needs; no AI system can perform this end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5This task has hard legal and ethical barriers: duty of care, liability for patient safety during transport and outings, licensing requirements for driving, and mandatory human supervision in healthcare contexts. Regulatory frameworks and professional standards require a qualified human to accompany vulnerable patients.
Adoption barriersclaude-sonnet-55/5Direct physical care, safety supervision, and liability for patient wellbeing during transport require a human aide; regulatory and safety requirements make this a hard barrier to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Providing safe transportation and accompaniment at quality requires either a human aide (current practice) or a fully autonomous vehicle with liability coverage (not yet economically viable). The all-in cost of current AI solutions would exceed that of an aide.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute for physical transportation and in-person supervision, so AI cost is not comparable—human labor is the only option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can reliably perform this task end-to-end; it requires embodied presence and judgment in dynamic physical and social contexts that current AI agents cannot handle. While autonomous vehicle research exists, it remains limited and regulatory barriers prevent widespread deployment.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product accompanies patients physically or drives them to outings; this remains entirely outside current product capability.

Related occupations — Healthcare Support

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.