Digital communication now plays a central role in how people collaborate, learn, and build relationships. However, in many real-world contexts—particularly in workplaces—communication is shaped by power dynamics and uneven social norms. Individuals in positions of authority may express themselves in ways that are brief, ambiguous, or overly direct, while recipients are often expected to interpret and respond appropriately. This asymmetry places a cognitive and emotional burden on readers, who must make sense of messages they cannot control, often leading to stress, negative affect, or misinterpretation.

Existing solutions largely adopt a writer-centric approach, focusing on helping authors refine tone, grammar, or wording. While useful, these approaches rely on the assumption that writers are willing or able to adjust their behavior—an assumption that frequently does not hold in practice, especially in hierarchical or high-pressure environments.

In contrast, this project proposes a reader-centric paradigm for digital well-being. By shifting the focus from expression to perception, this work repositions AI as a personalized mediation layer that empowers individuals to engage with digital communication in a healthier and more constructive way. Ultimately, the project aims to improve comprehension, reduce conflict, and support more balanced and resilient digital interactions.