Background: Artificial intelligence (AI) is rapidly transforming health care, yet little is known about how people living with arthritis perceive and interact with AI-driven tools. As these technologies expand in scope from AI scribes and symptom tracking to imaging interpretation and diagnostic support, it is important to understand patient readiness to engage with such tools.

Objectives: To examine how people living with arthritis currently engage with AI, their levels of trust and confidence, and their informational needs.

Methods: A national online survey was conducted by Arthritis Consumer Experts (ACE), a Canadian patient advocacy organization. The online survey was hosted on SurveyMonkey and open from June 18, 2025 to August 10, 2025 for people living with rheumatic conditions. The survey was primarily shared via ACE social media and email lists. The survey included closed- and open-ended questions which asked respondents about their current use, awareness of AI in specific clinical applications, level of trust, confidence in AI-generated recommendations, and gaps in knowledge regarding AI tools. Sociodemographic questions were also included. Quantitative responses were summarized descriptively, while qualitative comments were analyzed for key themes reflecting key concerns.

Results: A total of 282 individuals participated; 266 (94%) of whom identified as a patient living with rheumatic disease, 5 (2%) as a caregiver and 11 (4%) as both a patient and caregiver. Respondents were predominantly women (82%), highly educated (59% college/university graduates), living in urban centres with population of >100,000 (57%). The majority were diagnosed more than a decade ago, with 34% living with arthritis for >20 years. Six percent of respondents identified as Indigenous, 1% as Black, and 6% as People of Colour.

Familiarity with artificial intelligence varied widely. Although many participants (53%) reported they understood AI very well or fairly well, uncertainty regarding specific applications was common. When asked whether AI is currently being used for select purposes in arthritis care, 43% of respondents were unaware. Only a quarter (27%) of respondents reported using AI-powered tools in their arthritis care. Among those aware of AI applications, the most recognized were AI scribes (28%), diagnostic tools (21%), virtual assistants (23%), and symptom tracking tools (17%).

Trust in AI tools were mixed. While a majority of respondents reported at least moderate trust in AI-supported diagnosis (67%) and in AI’s ability to respond accurately to patient questions (65%), higher levels of trust were expressed for other applications. Specifically, more respondents reported at least moderate trust in AI to improve early detection of arthritis (73%), accelerate the development of new therapies (78%), identify causes of arthritis (78%), predict disease flares (73%), and improve access to arthritis care in underserved communities (75%). Others voiced concerns about loss of human oversight, lack of human connection, inaccurate health information and misdiagnosis, privacy concerns and environmental impact.

Confidence in AI-generated recommendations remained low, with respondents seeking assurance through scientific validation, clear accountability mechanisms, and lay-friendly education explaining how AI systems work. When asked what would allow them to feel more confident about the safety of AI-driven tools that are being used in arthritis care, respondents reported:

• Assurance from health care provider (72%)

• Research and science that supports the use of the AI-driven tools (71%)

• Transparency of the data set that is being used to train the AI (68%)

• Public policies and regulations to protect patients (63%)

• Easy-to-understand instructions (55%)

Conclusions: Our national survey demonstrates that while people living with arthritis are curious and reasonably optimistic about AI in arthritis care, substantial uncertainty persists. Patients want AI tools to work together with, not replace, clinician expertise, and many emphasized the need for accuracy and transparency. Accessible education and stronger communication about how AI is used in arthritis care is essential for building trust and confidence. As AI adoption accelerates, patient perspectives must be included in the development, regulation, and implementation of clinically meaningful AI-driven tools in rheumatology.

The findings are a call to action for governmental and institutional policy makers, pharmaceutical manufacturers, rheumatology researchers and clinicians, and patient organizations as AI increasingly becomes part of Canada’s healthcare present and future:

• Governments must ensure AI tools in arthritis care are designed with equity in mind and carefully regulate.

• At the development stage, people wanted reassurance that AI-generated drugs and recommendations are subject to human review and approval.

• Health care professionals should prepare for AI by integrating it into care in ways that support, not replace, their role.

REFERENCES: NIL.

Acknowledgments: NIL.

Disclosure of Interests: None declared.