Proactive Systems in HCI and AI: Concepts, Challenges, and Opportunities

The Problem of Proactivity

As AI and sensing technologies advance, systems are increasingly moving from reactive interfaces—which wait for explicit user commands—to proactive ones that anticipate needs and take initiative. While this shift promises greater efficiency and helpfulness, the term 'proactive' is currently used as a catch-all for diverse behaviors. It is often applied to simple recommendation engines or scheduled reminders, which lack the autonomous, goal-oriented, and future-focused characteristics of true proactive systems. This lack of a shared definition makes it difficult for researchers to compare systems, establish design best practices, or measure user outcomes effectively.

Limitations of Current Paradigms

Existing design and evaluation methodologies are largely rooted in reactive interaction paradigms. Standard usability metrics, such as the System Usability Scale (SUS), were designed for systems that respond to user input and fail to capture the unique challenges of proactive behavior. Specifically, these tools do not account for the critical nuances of autonomous action, such as the appropriateness of the timing, the user's sense of control, the transparency of the system's reasoning, and the calibration of trust. When a system acts without being asked, it introduces new risks regarding user consent and the perceived legitimacy of the system's initiative.

A Path Toward Standardization

The authors argue for a multidisciplinary effort to formalize the study of proactive systems. By drawing on organizational psychology—where proactivity is defined as self-initiated, future-oriented, and change-oriented behavior—the field can begin to distinguish between superficial system initiation and genuine proactive capability. The proposed workshop aims to bridge these gaps by developing a shared conceptualization, mapping the design space, and co-creating human-centered guidelines that prioritize user agency and trust in autonomous environments.