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

Shame and guilt are creating hidden AI use rather than stopping it: Students in the study engaged in "hiding behaviors and selective disclosure" of their AI tool use, suggesting that shame-based approaches may drive AI use underground rather than eliminate it or make it more responsible.

Emotional conflict persists alongside continued AI use: The research found that shame and guilt often coexist with ongoing AI use, creating "cycles of reduced agency and moral tension" rather than actually changing student behavior—indicating that emotional regulation alone is insufficient for promoting responsible AI practices.

AI use conflicts with students' professional identities: Students experienced tension between using AI tools and their self-image as "competent, hardworking, or ethical computing students," suggesting that how institutions frame AI use affects whether students see it as compatible with their values and career identity.

Current policies may be counterproductive: The authors note that "how AI policies are implemented and enforced" influences these emotional responses, implying that punitive or shame-based policy approaches could backfire by creating secrecy rather than fostering genuine responsibility.

Open dialogue is needed instead of stigma: The study recommends "classroom practices that can foster healthy, open discussion" around AI use, suggesting that transparent, non-judgmental conversations may be more effective than policies that trigger shame and hiding behaviors.