ADVICE TO THE PROFESSION: USING ARTIFICIAL INTELLIGENCE IN CLINICAL PRACTICE

Last Updated: August 2025

Artificial intelligence (AI) is when a machine or computer performs tasks that would typically require human intelligence. With AI, complex tasks, such as analyzing, reasoning, and learning, are performed without humans programming every step. These technologies are adaptive and keep learning and evolving as they receive new information, including in response to use in real-world settings.

Use and benefits of AI in health care

AI is rapidly evolving and is already contributing to improvements to health care and the way it is delivered. It has the potential to increase efficiency, support clinical decision-making, reduce costs, and alleviate some of the administrative burdens physicians face – allowing them to focus on direct patient care and improving access to care.
As the CMPA explains, AI can assist physicians with elements of care, and has the potential to increase diagnostic accuracy, improve treatment planning, and predict outcomes. AI has also shown potential in image-intensive fields, including radiology, pathology, ophthalmology, dermatology, and image-guided surgery, as well as in broader public health purposes, such as disease surveillance.

AI scribes

As part of efforts to support physicians and promote responsible digital health adoption, the use of AI scribes is encouraged to reduce time spent on routine administrative tasks and clinical documentation.
An AI scribe is an AI-powered tool that automatically captures and summarizes (and, in some cases, records) clinical encounters between a physician and a patient into detailed medical notes for the physician’s review in real-time.
AI scribes differ from dictation software which simply converts voice to text. An AI scribe can extract content from what it summarizes and apply it to appropriate fields of the patient’s medical record. Some AI scribes can analyze and isolate relevant medical information from the conversation between a physician and patient, and create a more concise medical note for the physician’s review.
An OntarioMD study found that AI scribes led to a 70 per cent reduction in time spent on paperwork, and over 3 hours per week saved doing administrative tasks. Primary care providers using AI scribes also reported reductions in cognitive load, and afterhours work; improved efficiency and documentation practices; increased job satisfaction, professional fulfillment, and work-life balance; and perceived improvements in quality of care.

AI tools for patient intake and triage

AI tools are transforming intake and triage processes, streamlining workflows, improving patient experiences, and helping to achieve more efficient use of health system resources.
AI intake tools can reduce time spent on routine administrative tasks by, for example, analyzing incoming communications received by clinics, categorizing them, and directing them to the appropriate staff or physician. These tools can also automatically contact patients to book follow-up appointments and send reminders about clinic appointments which can reduce patient no-shows.
AI tools can be used to triage patients by collecting and assessing patient data to rapidly sort and prioritize cases based on risk and urgency, and optimize patient queues in emergency departments. This can improve workflows and how patients move through the system, helping to enhance patient safety.

AI tools for supporting clinical decision making

AI is increasingly being used to support clinical decision-making by allowing physicians to input patient information to help the physician make a diagnosis, identify potential investigations, and manage the patient’s care.
AI tools are helping detect diseases earlier—sometimes even before symptoms appear. This may allow for quicker, more accurate diagnoses, and lead to earlier treatment and better outcomes. Some examples include colorectal cancer screening during colonoscopy to detect subtle lesions and AI review of CT coronary angiography to quantify vessel disease and predict future cardiovascular events.
AI is also supporting decisions during procedures such as laparoscopy and robotic surgery by assisting surgeons in real time to optimize their techniques during the surgery.

Role of AI

AI is meant to complement clinical care, not replace medical expertise. AI lacks human qualities such as creativity, intuition, and judgment that arise from years of experience. It is important that care always reflects a physician’s own recommendations based on evidence and good medical judgment.

Considerations when using AI in practice

When integrating AI applications into their practice, physicians need to be aware of the following:
- Accuracy: AI may sometimes produce false or inaccurate content. Physicians need to review all information generated using AI for accuracy and completeness.
- Accountability: physicians are ultimately accountable for their use of AI tools, including when using AI to support clinical decision-making or medical documentation (e.g., through AI scribes).
- Data privacy and protection: all patient data entered into AI applications needs to be kept private and secure. Physicians’ obligation to protect their patients’ personal health information is no different when using AI than in any other circumstance.
- Potential for bias: AI can repeat biases in the data it learns from, which can have adverse impacts on certain groups. Physicians need to ensure the AI tool they are using is appropriate for their patient population, and evaluate the output of the tool for bias.
- Transparency: physicians need to inform patients about how AI will be used, and in particular obtain patient consent before recording conversations using AI. physicians are ultimately accountable for their use of AI tools, including when using AI to support clinical decision-making or medical documentation (e.g., through AI scribes).
Physicians may want to consult the Canadian Medical Protective Association (CMPA) or the Ontario Medical Association (OMA) for further guidance on the use of AI.

Professional and legal obligations when using AI

Although there is not currently a specific law or policy addressing AI, the core expectations of physicians remain unchanged. Physicians who use AI are expected to be mindful of their legal and professional obligations, including privacy and confidentiality, how patient data will be transferred, stored, and used, and whether reasonable safeguards are in place to protect patient data.
In accordance with the Personal Health Information Protection Act, physicians are also required to obtain consent before collecting, using, or disclosing their patients’ personal health information. For more information, physicians can see the Protecting Personal Health Information policy.

Approved AI applications

CPSO cannot make recommendations about specific AI technology and applications. However, the Ministry of Health partnered with Supply Ontario and OntarioMD (OMD) to create an Ontario AI Scribe Program that features a list of Canadian AI scribe vendors that meet Ontario’s privacy and security standards. Please see the list of qualified vendors here, or visit the Ontario AI Scribe Program to learn more.
Canada Health Infoway also has an AI Scribe Program with a list of vendors that have been approved following a Pan-Canadian Procurement Process. See their website for more information on the program and their list of approved vendors.