Privacy is Not One-Click: Designing Robots That Adapt to Older Adults’ Changing Boundaries

Paper Title

Privacy is Not One-Click: Designing Robots That Adapt to Older Adults’ Changing Boundaries

Publication Info

• Topic area: Privacy design for social robots tailored to older adults
• Keywords: Social robots, older adults, privacy, participatory design, speculative design, multidimensional privacy, context-aware systems, data ownership, adaptive privacy, human-robot interaction

Background and Problem

• Problem / challenge: Current privacy mechanisms for social robots often rely on static, one-size-fits-all solutions that fail to address the dynamic, multidimensional nature of privacy as experienced by older adults.
• Significance: As social robots become increasingly integrated into older adults’ lives, addressing privacy concerns is critical to ensure trust, autonomy, and adoption of these technologies.
• Motivation and related work: Prior research has treated privacy dimensions (physical, informational, psychological, social) in isolation, neglecting their interdependence. Existing solutions like context-aware systems and transparency mechanisms lack holistic integration, leaving gaps in user agency and trust. This paper builds on these gaps by involving older adults in participatory design (PD) and speculative design (SD) to co-create privacy solutions.

Solution

• Proposed approach: A two-phase process combining participatory design workshops with older adults and speculative design by the research team to develop privacy-preserving features for social robots.
• Novelty:
  1. Empirical insights into how older adults conceptualize and negotiate privacy across four dimensions (physical, informational, psychological, social).
  2. A post-PD speculative design process to extend and refine privacy solutions based on older adults’ input.
  3. A set of multidimensional privacy design features, including context-aware privacy profiles, real-time feedback, and data ownership tools.
• Procedure and key techniques:
  1. Conducted two participatory design workshops with 14 older adults in assisted living facilities.
  2. Analyzed workshop outputs (storyboards, discussions) to identify privacy concerns and design ideas.
  3. Developed speculative storyboards to explore and refine privacy features, integrating older adults’ input with research expertise.
  4. Mapped proposed features to four privacy dimensions to ensure holistic coverage.

Results

• Concrete findings:
  • Older adults view privacy as dynamic and situational, requiring ongoing negotiation and control.
  • Key privacy concerns include skepticism toward cloud storage, preference for tangible controls, and distrust of static, one-time consent models.
  • Proposed design features include:
    1. Preset Privacy Profiles: Context-based modes (e.g., “Resting,” “Emergency”) with adjustable sensing and data-sharing settings.
    2. Real-Time Feedback: Visual, auditory, or mechanical cues indicating active sensors and data collection.
    3. Data Ownership Tools: Local storage via removable memory cards and event-based data logs for review and deletion.
• Advantage over baselines:
  • Moves beyond static, isolated privacy solutions by addressing privacy as a multidimensional, evolving process.
  • Empowers older adults with intuitive, tangible, and context-aware privacy controls.
• Experiments / evaluation:
  • Two workshops with 14 participants (7 female, 7 male) aged 65+.
  • Activities included robot demonstrations, storyboarding, and privacy scenario discussions.
  • Data analyzed through thematic and content analysis to extract privacy concerns and design ideas.
• Limitations and future work:
  • Limited participant diversity (Midwest U.S., assisted living residents).
  • No functional prototypes or long-term field studies; future work should include diverse cohorts and real-world testing of proposed features.

Summary

This study explores how older adults conceptualize privacy in interactions with social robots, emphasizing its dynamic, multidimensional nature across physical, informational, psychological, and social dimensions. Through participatory design workshops and speculative design, the authors developed privacy-preserving features, including context-aware profiles, real-time feedback, and data ownership tools. These features aim to empower older adults with intuitive and adaptable privacy controls. While the findings are based on a small, homogeneous sample, they provide a foundation for designing social robots that respect older adults’ evolving privacy needs and preferences.