Is your clinic ready for AI? A focus on consent and documentation 

Artificial intelligence (AI) is becoming increasingly common in clinical practice, from ambient scribes and documentation tools to clinical decision support. While these technologies can improve efficiency and reduce administrative burden, they don’t change a regulated member’s professional responsibilities, including protecting patient privacy and providing safe, high-quality care.

If your clinic is considering or already using AI, now is a good time to review your processes:

• Are staff and regulated members clear on when patients need to be informed?
• Does your documentation process reflect patient consent and how AI is being used?
• Have privacy requirements been addressed before introducing a new tool?

Build informed consent into your workflow

Patients should know when AI is being used as part of their care, whether it’s documenting an encounter, supporting clinical decision-making or performing another clinical function.

Meaningful informed consent means more than simply telling a patient AI is involved. Patients should have the opportunity to ask questions to ensure they understand what the technology is being used for and know they can decline its use as part of their care. A patient’s decision should be respected and the informed consent discussion, including the patient’s choice, should be documented in the patient record.

Clinics should also consider how they will respond if a patient does not consent to AI being used as part of their care. Establishing clear workflows in advance helps ensure patients continue to receive appropriate care while respecting their choice and supporting consistent documentation of the discussion and decision.

Staff who support patient intake or registration should also understand when questions about AI use may arise and where to direct patients for more information. CPSA’s Artificial Intelligence (AI) Advice to Albertans document may be a helpful resource to share with patients.

AI can support care, but it doesn’t replace professional judgment

AI can be a valuable clinical support tool, but regulated members remain responsible for every clinical decision and every entry in the patient record.

Before AI-generated content is finalized, it must be reviewed carefully for accuracy, completeness, bias and AI-generated inaccuracies. Clinic managers may wish to consider whether existing workflows allow sufficient time for this review before documentation becomes part of the health record.

Documentation matters

Patient records must accurately reflect the care provided, including when AI contributed to documenting the patient encounter, diagnosis or treatment decisions. Clear documentation supports continuity of care, promotes transparency and helps ensure the medical record remains a reliable account of the patient’s care.

Clinics should also remember that AI-generated information used in patient care forms is part of the patient record and remains accessible to patients.

Before implementing an AI tool

Successfully adopting AI involves more than choosing the right technology. Before introducing an AI tool, regulated members should ensure privacy requirements have been addressed, including completing or updating a Privacy Impact Assessment (PIA) where required under Alberta’s Health Information Act.

It’s also a good opportunity to review clinic policies, consent processes, staff education and documentation practices to ensure they’re aligned with legislative and regulatory expectations.

As AI technology continues to evolve, taking time to review policies, workflows and documentation practices can help support the safe and responsible use of these tools. Regardless of the technology being used, accountability for patient care always rests with the regulated member.

Questions? Review the full Artificial Intelligence in Clinical Practice Advice to the Profession for additional guidance and clarity. For additional questions, feel free to reach out to support@cpsa.ca for assistance.