AI for Occupational Therapists

Occupational therapy helps people do the activities that matter to them, from dressing and cooking to school and work. AI is showing up in four areas of that work. Documentation is the most common use. Ambient scribes and note-drafting tools can turn dictation or session audio into SOAP notes and evaluation reports. OTs still have to show skilled service. Payers such as Medicare expect documentation to explain why a licensed therapist's expertise was needed, not just list what the client did. AI drafts often describe the activities and leave out that reasoning. The COAST goal-writing framework (client, occupation, assist level, specific condition, timeline) gives AI a useful template for occupation-based goals. Activity planning is where generative AI is strongest. It can brainstorm themed activities, grade tasks up or down, write visual schedules, and adapt instructions for clients with cognitive impairments. The OT decides whether an activity fits the client's goals and is safe. Assistive technology has taken up AI quickly. Apps such as Seeing AI and the AI feature in Be My Eyes describe scenes and read text for people with low vision. Voice assistants control lights and thermostats, and word prediction and speech-to-text help with writing. OTs judge whether a tool fits the client, set it up, train the client and follow up, including teaching when not to trust the tool. Home assessments can use phone photos, video walkthroughs and measurement apps to gather information remotely. AI can help turn the findings into a report and list possible modifications. Judging fall risk, transfer ability and what a caregiver can manage still requires watching the person in the space. Two misconceptions are common. The first is that AI can score standardized assessments from video. Validated scoring still requires administration by a trained professional. The second is that a measurement app is precise enough for every decision. For critical clearances, confirm with a tape measure.