Background: In India, people living with axial spondyloarthritis (axSpA) face significant barriers to understanding their disease and treatment. Although English is an official language, many patients are more comfortable in regional languages and find medical reports, prescriptions, and clinical terminology difficult to interpret. This language barrier, combined with short and often rushed clinical consultations, limits patients’ understanding of disease progression, treatment rationale, and medication use, thereby reducing health literacy, self-advocacy, and meaningful participation in care. Limited understanding also makes it difficult for patients to explain their condition to family members. When information is available in the local language, family members are better able to read, understand the disease, and provide informed emotional, practical, and social support. AxSpA commonly affects individuals during their most productive working years. Pain, stiffness, delayed diagnosis, and reduced disease understanding can negatively impact education, employment continuity, career progression, and earning potential, increasing financial stress and reliance on family support.

Objectives: To strengthen advocacy, education, and awareness by using freely available Artificial Intelligence (AI) tools - particularly, large language models (LLMs) - to improve understanding of disease, medical reports, and prescriptions among people with axSpA through vernacular-language support, while also promoting work preparedness, digital inclusion, and socio-economic empowerment.

Methods: During counselling sessions with axSpA warriors, we collect socio-demographic information, including education level, financial status, and employment or educational background. Participants are connected to a patient-led digital ecosystem via WhatsApp, Telegram, and Facebook to support ongoing engagement, peer learning, advocacy, and mutual support. Patients are guided to share personal medical documents, such as blood tests, MRI and X-ray reports, and prescriptions, after masking personal identifiable information (e.g., name, date of birth, hospital identifiers), thereby promoting safe online practices. Documents are reviewed using LLMs that support vernacular languages. Using simple prompts - and, where needed, voice-based search and input - complex medical information is translated into plain, easy-to-understand language. Voice interaction improves accessibility for patients with limited typing skills, physical discomfort, or lower digital literacy. This approach enables patients to understand diagnosis, disease activity, progression, and medication purpose, and allows both patients and family members to read, revisit, and discuss explanations in their own language. Improved understanding helps patients prepare for medical consultations, frame questions in advance, and engage more confidently in shared decision-making. Many participants initially perceive AI as technical, intimidating, or inaccessible. We actively encourage exploration of AI to reduce fear and demonstrate its practical relevance. Through independent research, we systematically identify, evaluate, and curate learning materials from social media platforms, YouTube, free courses, and online portals, and use this knowledge to guide patients in creating effective prompts and applying AI meaningfully. Participants are also supported in using AI to improve CVs, identify transferable skills, explore flexible or remote work options, and access AI-enabled livelihood opportunities aligned with their physical limitations. All activities are volunteer-driven and unfunded. Despite limited resources, free learning materials are continuously identified and shared with motivated axSpA warriors. Planned next steps include structured Hindi and Marathi-language AI learning modules and the development of a peer mentorship network of axSpA warriors who have successfully transitioned to online or AI-supported work, further strengthening community-led empowerment.

Results: Participants reported a marked improvement in understanding medical reports and prescriptions when explanations were provided through LLMs in their local language. Vernacular-language interpretation reduced confusion related to medical jargon, improved awareness of disease progression and treatment intent, and enabled patients to explain their condition more clearly to family members, resulting in better-informed family support. Patients felt better prepared for medical consultations, more confident in asking questions, and more engaged in shared decision-making. Increased confidence in using AI - through both text and voice interaction—also translated into greater engagement with learning, CV enhancement, skill exploration, and identification of suitable employment or earning opportunities. Participants reported reduced fear of technology, improved digital confidence, and increased motivation to pursue flexible, remote, or AI-supported work, particularly among unemployed or home-bound individuals.

Conclusions: This unfunded, patient-led initiative demonstrates how LLMs can be leveraged as effective tools for advocacy, education, and awareness in axSpA care. Enabling patients and their families to understand disease, medical reports, and prescriptions in their own language strengthens patient voice, informed self-advocacy, family engagement, and meaningful participation in healthcare decisions. Beyond health literacy, this approach promotes safe digital practices, work preparedness, employability, and socio-economic empowerment by encouraging confident, guided use of AI. The model highlights the potential of community-driven, multilingual, and accessible AI adoption - including voice-based interaction - to reduce inequities, improve disease self-management, and enhance livelihood opportunities in resource-limited settings.

REFERENCES: NIL.

Acknowledgments: NIL.

Disclosure of Interests: None declared.