Co-designing MESA-Bot: Enhancing Accessibility, Privacy, Security, and Trust in a Mental Health Chatbot for Older Adults

Paper Title

Co-designing MESA-Bot: Enhancing Accessibility, Privacy, Security, and Trust in a Mental Health Chatbot for Older Adults

Publication Info

• Topic area: Mental health chatbot design for older adults, focusing on accessibility, privacy, security, and trust.
• Keywords: Mental health, chatbot, older adults, accessibility, privacy, security, trust, co-design, usability, digital health.

Background and Problem

• Problem / challenge: Older adults face barriers in adopting mental health chatbots due to accessibility, privacy, security, and trust concerns. Existing systems often fail to integrate these dimensions effectively.
• Significance: Addressing these barriers is crucial to making mental health support tools inclusive and effective for aging populations, who are particularly vulnerable to mental health challenges.
• Motivation and related work: Prior studies highlight the potential of chatbots to reduce loneliness and support mental health but reveal gaps in usability, accessibility, and trust for older adults. Privacy and security mechanisms are often treated as secondary, leaving older adults hesitant to adopt these tools.

Solution

• Proposed approach: MESA-Bot (Mental and Emotional Support Assistant for Older Adults), a co-designed, non-diagnostic mental health chatbot tailored to the needs of older adults.
• Novelty:
  1. Co-design process involving older adults to address cognitive, sensory, and trust-related barriers.
  2. Integration of transparent consent, simplified conversational flows, and adaptive emotional tone to enhance usability and trust.
  3. Implementation of privacy and security mechanisms, including revocable consent, role-based access control, and short data retention, validated through technical testing.
  4. Development of design principles for inclusive and trustworthy mental health chatbots.
• Procedure and key techniques:
  • Phase I: Co-design sessions with 10 older adults and analysis of 10 existing mental health chatbots to identify gaps and inform prototype design.
  • Phase II: Evaluation of the prototype with 28 older adults through semi-structured interviews and technical validation (e.g., STRIDE threat modeling, OWASP ASVS mapping, penetration testing, and red-teaming).

Results

• Concrete findings:
  • 86% of participants found MESA-Bot easy to use.
  • Transparent consent and privacy settings increased trust for 14 participants.
  • Key features like high-contrast interfaces, large buttons, and structured prompts were well-received.
  • Security mechanisms (e.g., consent gating, ephemeral data retention) were validated under adversarial testing.
• Advantage over baselines:
  • MESA-Bot addressed gaps in usability, accessibility, and trust identified in existing chatbots.
  • Privacy and security mechanisms were made transparent and actionable, unlike many commercial systems.
• Experiments / evaluation:
  • Phase I: Co-design with 10 participants to refine chatbot features.
  • Phase II: Evaluation with 28 participants, focusing on usability, trust, and privacy perceptions.
  • Technical validation through STRIDE threat modeling, OWASP ASVS mapping, and penetration testing.
• Limitations and future work:
  • Prototype lacks scalability for production use.
  • Limited inclusion of healthcare professionals in the design process.
  • Text-first interaction model excludes voice-based or multimodal interfaces, which will be explored in future work.

Summary

This study introduced MESA-Bot, a mental health chatbot co-designed with older adults to address barriers in accessibility, privacy, security, and trust. Through a two-phase process involving co-design and evaluation, the chatbot incorporated features like transparent consent, simplified interaction flows, and adaptive emotional tone, which were validated by 86% of participants as easy to use. Technical testing confirmed robust privacy and security mechanisms, including revocable consent and short data retention. While the prototype demonstrated strong usability and trust, future work will focus on scalability, voice-based interfaces, and integration of healthcare professionals' expertise. MESA-Bot offers a model for inclusive and secure mental health technologies for aging populations.