AI in Occupational Therapy: A Cautious Embrace
Artificial intelligence is rapidly transforming industries, and healthcare is no exception. From administrative automation to diagnostic support, the potential of AI is immense. But how is this technological wave being felt on the ground by practicing clinicians? An April-May 2025 survey of 278 Occupational Therapists (OTs) in Ohio offers a compelling snapshot of a profession cautiously navigating the new frontier of AI.
The findings reveal a community intrigued by the potential of AI but simultaneously held back by significant concerns about trust, accuracy, and a lack of formal guidance.
The Adoption Landscape: A Tale of Two Halves
The survey data paints a clear picture of a divided landscape. When it comes to professional work, AI adoption is still in its early stages. Nearly half of the OTs surveyed (46.4%) reported that they have not used AI for work-related tasks in the past six months.
However, this professional reluctance doesn't fully extend to personal use, where only 25.9% of OTs are non-users. This suggests a general familiarity with AI tools, but a hesitation to integrate them into clinical practice.
For the 53.6% of OTs who have adopted AI, usage is consistent. About one-fifth (19.7%) use it weekly or daily for work, indicating that once the barrier to entry is crossed, AI can become a regular part of the workflow.
The ROI of AI: Is It Saving Time?
For the OTs who have integrated AI into their practice, the benefits are tangible. AI users reported saving an average of 3.1 hours per week, with the median time saved being 2 hours. In a profession often burdened by administrative tasks, reclaiming this time can be a significant advantage, allowing for more focus on direct patient care, planning, and professional development.
AI in Practice: What Are OTs Actually Using It For?
The survey reveals that OTs are primarily leveraging AI as an informational and administrative assistant rather than a core clinical partner. The most common applications among AI users include:
● Answering work-related questions (37.6%)
● Generating easy-to-understand explanations of complex topics (37.6%)
● Generating emails (35.6%)
● Generating treatment goals (34.2%)
Notably, tasks at the heart of clinical documentation, such as writing session notes (12.8%) or full reports (23.5%), see lower rates of use. This pattern suggests that while OTs are comfortable using AI for brainstorming and communication, they remain the primary authors of core clinical records.
The Trust Deficit: A Barrier to Deeper Integration
Despite the time-saving benefits, a significant "trust deficit" appears to be the primary barrier to wider AI adoption. When asked if they trust AI, a mere 8.4% of OTs agreed or strongly agreed.
This skepticism is reflected in their perceptions of AI-generated content. While nearly a third believe AI can write reports (28.4%) or intervention goals (27.6%) as well as an average practitioner, a strong majority still prefer human-generated content for most clinical tasks.
The top concerns weighing on practitioners' minds are:
● Becoming over-reliant on AI (71.5% agree/strongly agree)
● Privacy and confidentiality breaches (65.5% agree/strongly agree)
● A loss of clinical skills (54.9% agree/strongly agree)
These anxieties are directly mirrored in the reasons AI users give for not using the technology more extensively, with "personal preference for human judgment" (49.0%) and "concerns about accuracy or reliability" (47.7%) topping the list.
The Path Forward: A Clear Need for Policy and Training
The survey highlights a critical gap in institutional support. 60.3% of OTs reported that their employer has no formal, written policies on the use of AI. This lack of guidance leaves practitioners to navigate complex ethical and legal questions on their own.
Furthermore, most OTs are learning about AI informally, primarily from colleagues (37.8%) and websites (26.3%). When asked what they need, the answer was clear: a strong desire for formal education, with 78.4% requesting webinars or online training and 45.0% wanting in-person training.
Final thoughts
The relationship between Occupational Therapists and artificial intelligence is still in its early stages. While many recognize its potential to streamline tasks and save valuable time, a deep-seated skepticism about its reliability, ethical implications, and impact on clinical skills remains.
For AI to move from a peripheral tool to a trusted clinical partner, the path forward must be paved with clear institutional policies, robust ethical guidelines, and accessible, high-quality training. By addressing these foundational needs, the field of occupational therapy can empower its practitioners to leverage the power of AI responsibly, ensuring that technology serves, rather than subverts, the ultimate goal of client-centered care.