AI literacy negatively moderates the trust–appropriate reliance relationship, with the trust slope becoming more negative as AI literacy increases¶

Claim · All claims
Evidence · 1 study (2 entries) · associational r2 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? The trust × AI literacy interaction was negative (b = -1.03, p = .014). → Griffin Pitts 2026 q3 i? At lower trust, higher AI literacy corresponded to higher predicted appropriate reliance, but this difference narrowed and slightly reversed at higher trust. → Griffin Pitts 2026 (2)

Evidence¶

Griffin Pitts 2026¶

Griffin Pitts, Neha Rani, and Weedguet Mildort. (2026). Trust and Reliance on AI in Education: AI Literacy and Need for Cognition as Moderators. Preprint, accepted to AIED 2026. https://arxiv.org/abs/2604.01114 doi:10.48550/arxiv.2604.01114

q3 · i? · associational · r2

Moderation regression in the Results section (Table 2) with standardized trust and AI literacy among 432 participants; the interaction coefficient was b = -1.03, p = .014. No standardized effect size is printed.

"The trust×AI literacy interaction was negative (b=−1.03, p=.014), meaning the estimated trust slope became more negative as AI literacy increased."

Griffin Pitts 2026 (2)¶

Griffin Pitts, Neha Rani, and Weedguet Mildort. (2026). Trust and Reliance on AI in Education: AI Literacy and Need for Cognition as Moderators. Preprint, accepted to AIED 2026. https://arxiv.org/abs/2604.01114

q3 · i? · associational · r2

The Results section describes the interaction pattern plotted in Figure 4a with 95% confidence bands; conditional slopes were approximately -3.59, -4.62, and -5.65 at -1 SD, the mean, and +1 SD AI literacy. No effect size is printed.

"At lower trust levels, higher AI literacy corresponded to higher predicted appropriate reliance, while this difference narrowed and slightly reversed at higher trust levels due to the steeper negative trust slope at higher AI literacy."

Discussion¶

Related Claims¶

• Programming self-efficacy and programming literacy did not significantly moderate the trust–appropriate reliance relationship — related
• Higher post-task trust in an AI assistant predicts lower appropriate reliance during AI-assisted programming problem-solving — related
• Need for cognition negatively moderates the trust–appropriate reliance relationship, with the trust slope becoming more negative as need for cognition increases — related
• The trust–appropriate reliance relationship is non-linear, with the decline attenuating at higher trust and no reversal within the observed data — related
• Perceived AI literacy moderates the direct path: the negative GenAI use–trust association is significant at low literacy and non-significant at high literacy — related
• AI literacy interventions recalibrate reliance on ML systems rather than simply increasing trust — a broader claim this one bears on
• The association between early reliance and negative impact strengthens as evaluation literacy increases — related