Exploring the Design Space of Privacy-Driven Adaptation Techniques for Future Augmented Reality Interfaces

Modern augmented reality (AR) devices with advanced display and sensing capabilities pose signifcant privacy risks to users and bystanders. While previous context-aware adaptations focused on usability and ergonomics, we explore the design space of privacydriven adaptations that allow users to meet their dynamic needs. These techniques ofer granular control over AR sensing capabilities across various AR input, output, and interaction modalities, aiming to minimize degradations to the user experience. Through an elicitation study with 10 AR researchers, we derive 62 privacy-focused adaptation techniques that preserve key AR functionalities and classify them into system-driven, user-driven, and mixed-initiative approaches to create an adaptation catalog. We also contribute a visualization tool that helps AR developers navigate the design space, validating its efectiveness in design workshops with six AR developers. Our fndings indicate that the tool allowed developers to discover new techniques, evaluate tradeofs, and make informed decisions that balance usability and privacy concerns in AR design.

• Human-centered computing → Scenario-based design; Mixed / augmented reality; • Security and privacy → Usability in security and privacy.