What it feels like to be an occupational therapist in an AI-driven world
Most days, your car doubles as a second office and your diary overflows with home visits, phone calls, notes and reports. As an occupational therapist living in this era, what keeps you going are the moments that actually change someone’s day, when a small bathroom change makes a shower safer, when a client stands with less pain and more confidence, or when your report helps a client access equipment or home modifications that support safer, more independent living at home.
At the same time, the world around your work is shifting fast. AI-polished graphics, perfect captions, slick videos and instant summaries now fill your feed; we even use AI to edit and enhance our own blog photos and Facebook posts. Used well, it makes some parts of our work easier and more efficient, so we can spend more time with our family after client visits, tons of paper works and managing our branding. It also raises an important question: how do we make sure technology supports the heart of occupational therapy, rather than quietly replacing it?
Human occupational therapy vs AI/robots: what can’t be automated?
It’s the way you walk into someone’s home and, within minutes, are building trust while you watch how they actually move, reach, transfer, organise and cope. You’re constantly grading the task, adapting the environment, reading family dynamics and choosing whether this is the moment to push, to pause, or just to listen. An algorithm can generate suggestions or templates, but it can’t sit at the edge of a bed and hold space while someone is terrified of falling, ashamed of needing help, or grieving the loss of a role. It won’t notice the way a client hesitates before attempting a step, the strain in a carer’s voice, or how pain, fatigue, history and identity all collide in one simple activity of daily living and respond in real time the way a human occupational therapist can.
Here’s a practical way to think about it:
| OT task | What humans (OTs) do best | What AI does best | Likely future |
|---|---|---|---|
| Functional assessment (FCA / initial assessment) | Build rapport in the home, read body language and tone, weigh risk and safety, synthesise history, environment and performance into clinical judgement that informs NDIS/SAH recommendations. | Pull together past notes, outcome measures and device data; summarise patterns over time; flag possible risks or changes for the OT to review. | Human-led, AI-assisted |
| Goal setting & care planning | Use coaching skills, motivational interviewing and shared decision-making to uncover what actually matters to the person and their family, and co-design goals that are meaningful, realistic and aligned with NDIS/SAH plans. | Draft goal options and sentence structures (e.g. SMART goals) from prompts; organise goals under domains; check consistency across documents. | Human-led |
| Home modifications (NDIS / SAH) | Observe real-world task performance, weigh risk vs independence, consider cultural and family context, integrate building constraints, quotes and client preferences into defensible recommendations. | Search and compare fixtures/equipment, collate supplier specs, organise quotes, generate draft comparison tables and basic diagrams for reports. | Shared (OT clinical reasoning + AI research/admin support) |
| Therapy delivery & review | Adapt on the fly during sessions, grade activities, co-regulate distress, respond to pain/fatigue, educate families, and adjust the plan based on how the person is actually coping in their real environment. | Send reminders, track practice frequency, provide simple exercise prompts or education snippets, visualise progress data between OT reviews. | Human-led, tech-supported |
| Documentation & reports | Apply clinical nuance, ethics and accountability; link observations to NDIS/SAH criteria; justify recommendations clearly and fairly; ensure accuracy and professional responsibility. | Draft outlines, restructure sections, transcribe notes, suggest wording options, check consistency across long reports and templates. | AI-supported (OT edited and accountable) |
Professional bodies and health leaders are increasingly talking about technology as augmentation, not substitution. That’s because real OT outcomes still depend on human-centred reasoning, collaboration, and real-world adaptability, especially in complex systems like the NDIS and Support at Home.
How AI can realistically help occupational therapists
AI is most powerful when it takes the busywork and strengthens the data layer, so you can spend more time on clinical thinking, family conversations and actual therapy.
1. Analysing client data (wearables, routines, environment)
AI tools can help interpret trends from heart rate, sleep patterns, step counts, fatigue ratings, activity logs or even smart-home sensors. Over time, they may flag possible issues like reduced activity levels, disrupted sleep, or patterns that suggest increased falls risk, so you can review and decide whether earlier intervention is needed. The OT still decides what this means in the context of the person, their diagnosis and their goals.
2. Smarter assessments and progress tracking
Digital platforms can auto-score some tasks (e.g. cognitive training apps, handwriting or fine-motor programs, reaction-time tests) and highlight change from baseline. Used well, this can give you an extra layer of objective data to sit alongside your observation, interview and standardised assessments. It’s additional information, not a replacement for functional assessment, clinical reasoning or NDIS/SAH decision-making.
3. Streamlining documentation and report-writing
For many OTs, this is the first “easy win.” Secure AI scribes, speech-to-text and drafting tools can reduce the time you spend on notes, summaries and formatting especially for long NDIS functional capacity assessments or home modification reports. The non-negotiable: you still review, edit and ensure the language is accurate, person-centred and aligned with professional and AHPRA standards.
4. Supporting clients between sessions
Think of AI as a guided prompt, not a therapist. For example:
- A virtual assistant that steps a brain injury client through their morning routine.
- Pacing reminders for someone living with chronic pain or fatigue.
- Brief, app-based breathing or grounding prompts for anxiety.
These tools can extend your interventions between visits. As long as you carefully match the tool to the person, explain how it works, and regularly review whether it’s helping or overwhelming.
5. Smart-home independence plans (NDIS / SAH)
AI-enabled environments can include fall-detection, voice-controlled lights and appliances, routine prompts, medication reminders and safety alerts. For NDIS and Support at Home participants, this can support independence and carer sustainability. OTs are uniquely placed to assess suitability, prescribe options, and configure these systems. They support the person’s goals and cognitive, sensory and physical capacities rather than adding complexity or stress.
6. Robotics as “assistive technology 2.0”
Rehabilitation robots, robotic arms and wearable exoskeletons are beginning to appear more in rehab and community practice. They can increase repetition, support intensive practice, or allow more independent exercise with remote monitoring. Your role doesn’t disappear, it shifts into selection, safety, setup, training, grading and ongoing review. The technology delivers movement; you ensure it serves meaningful occupations, risk is managed, and goals remain person-centred.
Real concerns about AI in occupational therapy
The worries are not overreacting and occupational therapists should be loud, practical voices in how AI is used.
- Client data privacy and security
Wearables, apps, smart homes and AI scribes collect highly sensitive health and behavioural data. For Australian OTs under AHPRA, NDIS and Support at Home, that means clear consent, clear explanation of what’s collected and why, and safe storage and sharing. If you couldn’t comfortably explain the “data trail” to a client or carer, something needs to change. - Bias in algorithms
AI trained on narrow data can misread disability, culture, language, neurodiversity, mental health or socio-economic context. The result can be unfair risk flags or poor-fit recommendations. Human oversight, cultural humility and critical thinking stay essential. - Over-reliance and “automation creep”
If clinicians lean on auto-summaries more than client stories, we risk losing occupational narratives, the meaning and identity behind a task. AI can draft, but only you can listen deeply and decide what truly matters for this person in this environment. - Equity and the digital divide
Many NDIS participants and older adults in SAH lack reliable internet, modern devices or tech confidence. Tech-heavy solutions can unintentionally widen gaps, especially in rural or low-income communities. OTs should advocate for technology that actually reduces barriers. - Liability, scope and clinical governance
If an AI-generated suggestion is unsafe or inaccurate, the registered clinician is still accountable for what goes in the record and what is recommended. Clear policies, supervision, documented reasoning and cautious use of AI tools protect both clients and therapists.
So… will AI and robots replace occupational therapy?
Let’s look at what we already know. Automation-risk tools like Replaced by Robot estimate around a 0.35% chance of automation for occupational therapists, ranking OT about 6th out of more than 700 jobs as one of the least likely to be replaced. Another source, Will Robots Take My Job?, lists OT as minimal risk (0.0%). Professional bodies such as AOTA echo this message, describing AI as supporting information that should inform, not replace, human clinical decisions.
Put simply: AI is far more likely to change how we work than to remove why OT exists. And that’s actually the exciting part because as technology grows, someone still has to translate tools into real homes, real routines and real lives. That “someone” is an occupational therapist.
OT will evolve
If you feel uneasy about AI and robots, you’re not alone. It’s moving quickly, and it’s normal to wonder where you fit as the landscape shifts.
But the heart of occupational therapy remains the same: helping people do what matters, in a way that fits their life, their culture and their goals. AI may help you spot patterns sooner, write notes faster, or trial smarter assistive technology. It may tidy up your reports or support practice between sessions.
What it can’t replace is your presence in the room, your ability to listen, your ethical judgement, your creativity, and the way you make therapy meaningful in real homes and real communities.
And perhaps that’s the real opportunity: the better the tools become, the more important it is to have occupational therapists guiding them with humanity, critical thinking and care.
Thank you for reading! Blog written by Daniel Castrucci, Occupational Therapist and CEO of Cas Health.
reference:
https://www.replacedbyrobot.info/18537/occupational-therapist-ot
https://www.aota.org/publications/student-articles/career-advice/ai-innovation-instability
https://www.uscareerinstitute.edu/blog/65-jobs-with-the-lowest-risk-of-automation-by-ai-and-robots
Disclaimer: This blog is for general information and education only. This does not replace independent clinical judgement, local policies or funding rules. Any sections not specifically referenced or cited reflect the professional opinion and clinical experience of Daniel Castrucci at the time of writing. Technology, guidelines and program criteria (including NDIS and Support at Home) change over time, so readers should confirm details with official sources and seek personalised advice for their own situation.
Explore related blogs:
