AI impact · verified capabilities
How AI is changing the work of Physical Therapists
Our reviewed registry links 7 verified AI capabilities to tasks that are part of this occupation. This shows where AI tools can already assist with specific tasks. It is not a forecast that the occupation will disappear, and it does not measure how widely any tool is used.
Last reviewed: 2026-09-27 · O*NET 29-1123.00 · Mapping confidence: high
Task areas where verified capabilities connect
- Patient care
- Clinical documentation
- Diagnostics and imaging
- Hospitality service
- Assessment and review
- Field and technical work
Verified capabilities (up to 5)
- EmergingEvidence: Strong
Clinical record summaries
What AI can assist with: Extracts patient information and summarizes records over time.
Practical meaning: this can speed up part of the task, but still needs human direction and review.
Human responsibility: you review every output and keep the final decision.
Source: AWS - EmergingEvidence: Strong
Real-time voice agents
What AI can assist with: Holds spoken conversations in real time for defined service or sales flows.
Practical meaning: this can execute a bounded workflow, but ownership, controls and edge cases remain with people.
Human responsibility: you set permissions and approval points and stay accountable for outcomes.
Source: Microsoft - EmergingEvidence: Strong
Medical coding assistance
What AI can assist with: Uses clinical language processing to suggest codes and abstract records.
Practical meaning: this can speed up part of the task, but still needs human direction and review.
Human responsibility: you review every output and keep the final decision.
Source: AWS - NowEvidence: Strong
Biomolecular interaction prediction
What AI can assist with: Predicts molecular structures and interactions beyond single proteins.
Practical meaning: this can speed up part of the task, but still needs human direction and review.
Human responsibility: you review every output and keep the final decision.
Source: Google DeepMind - NowEvidence: Strong
Machine-learning sepsis risk flagging
What AI can assist with: Clinical models continuously or periodically analyze patient data to flag elevated sepsis risk for healthcare-provider assessment.
Practical meaning: this can speed up part of the task, but still needs human direction and review.
Human responsibility: you review every output and keep the final decision.
Source: U.S. Food and Drug Administration
More source-verified AI capabilities
Additional reviewed research for this occupation. These do not replace the core capability list or prove real-world adoption.
Conversational AI rehabilitation coaching
Sword Health documents Phoenix as a conversational AI that guides members during movement, adapts to session performance, and summarizes insights for clinicians.
Evidence: Limited
Practical meaning: a defined part of this workflow can be automated; judgement and exceptions remain human work.
Verified sourceAI-assisted patient instruction drafting
Nabla documents automatic generation of structured patient instructions from clinical conversations.
Evidence: Limited
Practical meaning: a defined part of this workflow can be automated; judgement and exceptions remain human work.
Verified sourceAI-assisted rehabilitation movement analysis
Hinge Health documents TrueMotion computer vision that tracks full-body movement, gives real-time feedback, and provides data insights physical therapists use to monitor and adjust care plans.
Evidence: Limited
Practical meaning: this can speed up part of the task, but still needs human direction and review.
Verified source
What remains human responsibility
Evidence shows a capability exists; it does not prove adoption in every workplace or full-task replacement.
Evidence limits
Each capability is verified from the cited source only. A source shows that a capability exists; it does not establish adoption, deployment depth or outcomes in any particular workplace.
Sources
- AWS — Generative AI and NLP for healthcare and life sciences
- Microsoft — Transparency note for real-time voice agents in Dynamics 365 Contact Center
- Microsoft — Computer-using agents, workflows, and real-time voice
- Google Cloud — Gemini Live API on Vertex AI
- Google DeepMind — AlphaFold
- U.S. Food and Drug Administration — 510(k) Premarket Notification: Bayesian Health Sepsis Flagging Device
- npj Digital Medicine — An artificial intelligence-powered learning health system to improve sepsis detection and quality of care: a before-and-after study
- Internal and Emergency Medicine — Real-time machine learning-assisted sepsis alert enhances the timeliness of antibiotic administration and diagnostic accuracy in emergency department patients with sepsis: a cluster-randomized trial
- Sword Health — Conversational AI rehabilitation coaching
- Nabla — AI-assisted patient instruction drafting
- Hinge Health — AI-assisted rehabilitation movement analysis
See what this means for your own week
The free assessment looks at the tasks you actually do and your own AI habits.
Start the free assessment