AI content Is not neutral
Why Sjögren’s is left out
Artificial intelligence is now deeply embedded in how health information is created and shared. From symptom explanations and blog articles to search engine summaries and patient education tools, AI-generated content is increasingly shaping how people understand illness.
While AI has the potential to improve access to information, it is not neutral. The quality and accuracy of AI-generated health content depend entirely on the data it is trained on. For autoimmune and under-recognised diseases such as Sjögren’s, this creates a serious risk of continued misrepresentation, minimisation, and exclusion.
The rise of AI-generated health information
Many people now encounter health information through AI-powered tools, whether knowingly or not. Search engines summarise medical conditions using AI. Websites generate articles with AI assistance. Some people even ask AI tools directly about symptoms before seeing a doctor.
This shift matters because AI-generated information is often perceived as authoritative. When information is delivered confidently and clearly, it can feel trustworthy, even when it is incomplete or outdated.
For conditions that are already poorly understood, this confidence can cause harm.
How AI learns and why that matters
AI systems do not independently research diseases. They learn patterns from existing data such as published research, medical textbooks, websites, and online discussions. If a condition is under-researched, inconsistently described, or historically dismissed, that bias is reflected in the output.
Sjögren’s is a clear example. Despite being a systemic autoimmune disease that can affect multiple organs, it is still commonly described in simplified terms. Much of the publicly available content focuses on dry eyes and dry mouth, while fatigue, pain, neurological involvement, and organ complications receive far less attention.
When AI is trained on this limited narrative, it reproduces it.
How Sjögren’s is commonly misrepresented by AI
AI-generated content about Sjögren’s often reflects the same issues patients encounter in clinical settings.
The disease is frequently reduced to a dryness condition rather than recognised as a systemic autoimmune disease.
Extraglandular symptoms such as severe fatigue, joint pain, neuropathy, lung involvement, and cognitive dysfunction are often missing or briefly mentioned without context.
Seronegative Sjögren’s is frequently overlooked, reinforcing the false idea that blood tests alone rule the disease in or out.
The lived experience of patients is rarely reflected, meaning the daily impact of the disease is minimised.
- This creates a narrow and inaccurate picture that does not align with how Sjögren’s affects real people.
Why this matters for patients
Many people with Sjögren’s experience long delays before diagnosis. Some spend years being told their symptoms are unrelated, stress-related, or not significant.
When patients turn to AI-generated content and do not see their symptoms reflected, it can reinforce doubt. People may question whether their experiences are valid or serious enough to pursue further care.
For newly diagnosed patients, inaccurate or incomplete information can also create confusion and fear. When fatigue, pain, or neurological symptoms are not acknowledged, patients may feel unprepared for what lies ahead.
AI content that omits or minimises symptoms contributes to the same pattern of invisibility that people with Sjögren’s already face.
Risks for clinicians and health systems
AI tools are increasingly being explored in healthcare education, triage systems, and clinical decision support. While these tools may improve efficiency, they also carry risk when the underlying data is incomplete.
If AI-generated summaries reinforce outdated understandings of Sjögren’s, they can influence clinical assumptions. This is particularly concerning for diseases that already suffer from delayed diagnosis and inconsistent recognition.
Technology that reflects existing gaps in knowledge does not solve the problem. It amplifies it.
The role of patient organisations in the AI era
As AI becomes more prominent, trusted patient organisations play a critical role in shaping accurate narratives.
Organisations like Sjogrens Australia provide content that reflects current research, clinical complexity, and lived experience. This information is not only valuable for patients and families but also contributes to the broader information ecosystem that AI systems draw from.
When patient-led, evidence-based content exists, it increases the likelihood that future AI-generated information will be more accurate, nuanced, and representative.
Advocacy, education, and awareness work do not just support individuals. They help ensure that emerging technologies do not leave rare and autoimmune diseases further behind.
A responsible way forward
AI can be a useful tool when used responsibly. It can improve access to information, support health literacy, and assist with education. However, it must not replace clinical expertise, patient experience, or critical thinking.
For autoimmune diseases like Sjögren’s, accuracy matters. Representation matters. Context matters.
AI should be treated as one source of information, not the final authority. Patients, clinicians, and policymakers must continue to rely on peer-reviewed research, specialist input, and trusted patient organisations.
Technology should support better understanding, not reinforce old misconceptions.
AI has the power to amplify awareness or amplify neglect. For Sjögren’s and other under-recognised autoimmune diseases, the difference lies in whose voices are included and whose knowledge is valued.
As AI continues to shape health information, ensuring accurate, comprehensive, and patient-centred content is not optional. It is essential.
References: World Health Organization. Ethics and governance of artificial intelligence for health. https://www.who.int/publications/i/item/9789240029200 National Institutes of Health. Sjögren’s Disease Overview. https://www.niams.nih.gov/health-topics/sjogrens-disease Baimpa E, Dahabreh IJ, Voulgarelis M, Moutsopoulos HM. Hematologic manifestations and predictors of lymphoma development in primary Sjögren syndrome. Journal of Autoimmunity. Ramos-Casals M, Brito-Zerón P, Bombardieri S, et al. EULAR recommendations for the management of Sjögren’s syndrome. Annals of the Rheumatic Diseases. Topol E. High-performance medicine: the convergence of human and artificial intelligence. Nature Medicine.
