Contradictions and Boundary Negotiation in Human–AI Interaction: Insights into Cyborg Anxiety and AI Literacy Among University Students

Generative artificial intelligence (GenAI) has rapidly entered university students’ academic and everyday lives, yet its implications for human–computer interaction remain underexplored. Using a qualitative diary study, this research analyzed 141 diary entries from undergraduate communication students in Eastern China (June 5–25, 2025) through reflexive thematic analysis. Findings reveal a three-stage mechanism shaping student experience: Contradiction → Negotiation → Subjectivity Outcomes. Students described a persistent tension between functional dependence on GenAI for efficiency and perceived erosion of subjectivity, including cognitive atrophy and self-worth doubt, often experienced as a fear of displacement. In response, they engaged in boundary negotiation through contextual role allocation and the creation of forbidden zones in areas such as original creation and emotional expression. This process produced two outcomes: Cyborg anxiety and critical boundary literacy. The study shifts human–AI research from adoption to identity and existence.