Stuck in a Spiral": Shame and Guilt as Social Regulators of AI Use in Computing Education

Synthesis: An interview study with 19 computing students through a functionalist perspective of shame and guilt. Findings show these emotions regulate when and how students make their AI use visible, engaging in hiding behaviors and selective disclosure. Students described shaming themselves, peers, and faculty for using AI. Shame and guilt coexist with continued AI use, creating cycles of reduced agency and moral tension rather than promoting behavior change. Students used language and metaphors of addiction. Highlights need to consider socio-emotional aspects of AI use in policy and classroom practice.

Key Findings

• Shame and guilt regulate when and how students make their AI use visible, producing hiding behaviors and selective disclosure.
• Students described shaming themselves, peers, and faculty for using AI, indicating that these emotions circulate socially rather than only privately.
• Shame and guilt coexist with continued AI use, creating cycles of reduced agency and moral tension rather than promoting behavior change.
• Students used language and metaphors of addiction to describe their relationships with AI use.
• These experiences emerged alongside broader feelings of nihilism and pessimism about their academic futures.

Study Design & Method

The study interviewed 19 computing students and analyzed their relationships with AI use through a functionalist lens of shame and guilt — that is, asking what social work these emotions do in regulating behavior. The analysis examined how students' emotional experiences relate to academic identity, confidence, and self-perception, and how those experiences intersect with institutional responses such as surveillance and detection.

What this means for practice

• Instructors. Replace detection-first messaging with classroom conditions in which AI use can be discussed candidly, since shame led students to hide their use rather than change it.
• Instructors. Let peers lead or co-lead these conversations, or train the staff who run them to avoid imposing normative judgments; this study used trained peer interviewers precisely because being questioned by faculty skewed disclosure.
• Administrators. Audit surveillance and detection policies for socio-emotional cost, because students described shaming themselves, peers and faculty for AI use, so enforcement can deepen the spiral of reduced agency and moral tension rather than reduce the behavior.
• Instructors. Treat addiction-like language and nihilism about academic futures as signals that a student's relationship with AI needs support rather than sanction, since the emotional valence of AI use is itself consequential for persistence in computing.

Limitations

• The study is a qualitative, exploratory investigation of 19 semi-structured interviews with computing students in North America, intended to surface experience rather than test hypotheses; the authors state the sample is not representative of all students, cultures or institutional contexts.
• All accounts are self-reported, and the authors note that because shame and guilt are sensitive, participants may have underreported, reframed or unevenly disclosed their behavior despite assurances of confidentiality.
• The study deliberately collected no academic performance data, which the authors state limits any examination of how shame, guilt or AI reliance vary with students' academic standing.
• Interviews were held by trained peer interviewers to mitigate power dynamics such as being questioned by faculty, so the accounts are participants' interpretations reported as illustrative and preliminary, and the functionalist lens may not capture identity judgments such as being "not real programmers."

Citation

Hamilton, K., Hou, I., Patel, D., Nnam, S., Patel, H., & MacNeil, S. (2026). "Stuck in a Spiral": Shame and Guilt as Social Regulators of AI Use in Computing Education.