Description This concept paper introduces the AI Multi-Card, a card-shaped, privacy-respecting AI companion designed as a deliberate alternative to always-on AI wearables such as smart glasses. The AI Multi-Card focuses on intentional interaction instead of continuous observation, enabling identification, authorization, verification, and context-specific assistance without permanent sensing or passive data collection. The device is conceived as a tool-like AI interface: activated only when explicitly required, physically transparent in its operation, and compatible with strict privacy and regulatory environments. Its proposed hardware architecture includes a small display, a miniature camera optimized for QR codes and documents, secure cryptographic elements, and low-power edge AI capabilities. Continuous video streaming and passive monitoring are explicitly excluded by design. The paper discusses the motivation behind this approach, outlines the physical and technical architecture, and analyzes its functional scope, including digital identity presentation, offline authentication, access control, and AI-assisted guidance. Particular emphasis is placed on privacy by design, user sovereignty, and institutional compatibility (e.g., public administration, healthcare, education, and transportation). By separating contextual AI assistance from permanent body-worn perception, the AI Multi-Card proposes a scalable and socially acceptable path toward everyday AI integration. The concept is intended as an open, extensible foundation for democratic, human-centered AI systems and future digital identity infrastructures. The work is published under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license to encourage reuse, adaptation, and further research.
Robert Alexander Massinger (Fri,) studied this question.