AI is already transforming healthcare, providing administrative efficiencies and reduced cognitive load for providers and streamlining health information for patients at home. The presence of this powerful yet imperfect technology cannot be denied. Throughout my 13 years of clinical experience in physical therapy (PT), I’ve watched AI move from the periphery of our work up to the edges of our clinical reasoning. We should determine as physical therapy clinicians not whether AI will find its way into our clinics, but how.
This shift can feel uncomfortable, but it also offers us an opportunity to adopt AI responsibly so that it augments the strengths of physical therapy clinicians and fills in known gaps, rather than replacing valuable experience and expertise. Technology that used to be reserved for administrative tasks can now help us surface evidence faster, spot clinical patterns sooner, and extend patient care beyond the clinic.
The true promise of AI lies not in autonomy of either the clinician or machine, but in a balanced collaboration that bolsters the strengths of each.
Maximizing patient care
Providing the very best care to our patients is our highest priority as clinicians. It’s our responsibility to make the most of reliable technology that is available to us and remain open minded to new advancements.
We’ve long known that lack of patient adherence is one of the primary contributors to poor patient outcomes with physical therapy. There are many reasons a patient may struggle with this: uncertainty they’re doing the exercises correctly, difficulty fitting a PT routine into their daily schedule, or a lack of understanding of how the program will help their condition, to name a few. Patient engagement with home programs is a powerful example of something that can be meaningfully supported by AI.
Computer vision technology can be used to assist patients while they’re performing their home exercises. The goal of this technology is not just telling someone that they’re moving wrong – there’s more to it than simply correcting form. As PTs, we want to encourage movement by using this technology to teach patients new exercises and engage with patients while they’re moving, and gathering valuable patient feedback. This technology can provide information about movement modifications when a patient is struggling or recommend progressions for the patient’s physical therapist to review if they think an exercise is too easy, rather than the patient waiting until their next appointment. Then, valuable insights can be delivered back to the physical therapist who remains the primary decision maker for that patient’s care journey.
Using AI agents or chatbots may also help extend PT care beyond the clinic. These tools can be built to engage patients within their home program, supporting patients on habit-building skills to help them fit exercise into their schedule. Educational components to a treatment plan, like pain neuroscience education, can be reiterated through compassionate conversation. Again, the goal of these tools is not to replace the provider. With careful implementation, these tools will engage with patients while also gathering information that can be summarized for a PT to easily determine whether modifications to a treatment plan are necessary.
The bottom line? It’s responsible and necessary to thoughtfully use AI that supports adherence and engagement under PT oversight.
Enhancing clinical decision making
Clinical decision making is not purely mechanical or logistical, but interpretive, contextual, and relational. The potential of AI arises from strengthening clinician intuition with reliable, reproducible input. Human clinicians can vary widely in their judgments and are prone to bias. Inter- and intra-rater variability is well documented in physical therapy assessments including joint palpation, movement and postural analysis, and range of motion measurement.
For example, a 2016 study published in JOSPT demonstrated anchoring bias by physical therapists measuring wrist passive range of motion when they were given different historical information about the patient prior to taking the measurement. AI algorithms may help reduce this variability by providing stable baselines and highlighting discrepancies.
We must strive to achieve synergy between AI and human clinicians, meaning that the best capabilities of each complement one another, becoming even greater than the sum of its parts. To achieve this, any tool that is created to support clinical work must also involve expert clinical input in development, from the brainstorming process through implementation, daily use, and regular iteration. If practicing clinicians perceive AI as opaque, intrusive, or misaligned with their goals, they will rightly disengage or override it reflexively. The solution is a dynamic equilibrium in which clinicians neither dismiss nor defer to AI, but engage with it as a thinking partner.
New research can take decades to become standardized across clinical care. AI can help these new standards become commonplace by summarizing bodies of evidence and surfacing important research to clinicians quickly. However, many of these tools cannot yet consistently assess the quality of sources or study design. The onus of determining the quality and reliability of these sources still rests solely with clinicians.
Remember: decision quality improves when reliable, reproducible inputs meet expert clinical interpretation.
Meeting patients where they are
The New York Times recently cited a survey from health policy research group KFF that found “Last year, about one in six adults – and about a quarter of adults under 30 – used chatbots to find health information at least once a month.”
Today, the world of advice for people in pain is so noisy, with countless online voices offering recommendations that can end up ultimately worsening their symptoms. Right now, those online recommendations are louder than some of the more evidence-based and optimistic advice around pain and movement, which can result in AI models surfacing less reliable information for people who are eager for relief. A 2025 study examined the performance of several widely used LLMs and found high variability and inconsistent accuracy compared to published clinical practice guidelines for lumbosacral radicular pain.
Because it’s no secret that patients are already using LLMs like ChatGPT to learn about their conditions, I invite them to bring what they find to our sessions. This presents a valuable opportunity to explain how I combined their story, exam findings, imaging when relevant, and their goals with the best available evidence to build the plan. This method allows us as the experts to coach patients toward safer self-education and correct misinformation in real time.
After all, AI should be seen as a bridge, not a barrier. These teachable moments build health literacy and keep care grounded in guidelines and context.
An optimistic future with thoughtful design
The promise of AI in the PT clinic is reallocating clinicians’ cognitive energy toward the type of thinking that is uniquely human: interpreting data with a lens that includes clinical history, patient beliefs, and experience gained through years of treating patients.
I’m energized by AI’s potential when used responsibly to shorten the distance between evidence and practice, to support newer clinicians in building robust clinical reasoning, and to give us earlier signals when a plan may need to change. At its best, AI can propel our potential as clinicians to new heights, as long as we set guardrails, measure what matters, and keep our identity as movement experts who combine science, skill, and relationship.
Photo: Irina_Strelnikova, Getty Images
Dr. Claire Morrow is a Doctor of Physical Therapy and Head of Clinical Consulting at Hinge Health. Her work includes supporting members, training physical therapists on evidence-based practices, advising product leadership, and contributing to commercial initiatives. She serves as the physical therapist representative on the Hinge Health clinical leadership team. Dr. Morrow is a board-certified Orthopedic Clinical Specialist and a Fellow of the American Academy of Orthopaedic Manual Physical Therapy. She is also clinical faculty for the Kaiser Permanente Northern California Orthopaedic Physical Therapy Residency, mentoring newly licensed physical therapists. Her prior experience includes outpatient care in hospital-based clinics, onsite employer clinics, and working with professional athletes. Her professional mission is to empower individuals to better understand pain, improve function, and enhance quality of life.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
