Longitudinal studies of trust and reliance in AI advice examine whether users rely appropriately over time but not whether independent reasoning capacity is maintained¶

Claim · All claims
Evidence · 1 study · theoretical r2 · q1
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? Existing longitudinal work shows trust is dynamic, shaped by AI accuracy, explanation type, and the timing of errors, but focuses on reliance rather than maintained reasoning capacity. → Sander de Jong 2026

Evidence¶

Sander de Jong 2026¶

Sander de Jong. (2026). Open Questions Towards Skill-Sustaining Reliance in Reflective AI Engagement. Workshop paper presented at CHIWORK'26 Workshop on Human Agency and Skill in AI-Supported Work. https://arxiv.org/abs/2609.27726 doi:10.48550/arxiv.2609.27726

q1 · i? · theoretical · r2

Theoretical argument in Section 5 reviewing longitudinal studies: Kahr et al. on trust development in repeated legal decision-making, Riedl and Bogert on chess players' use of AI feedback, and Kahr et al.'s field study with logistics professionals.

"they focus on whether users appropriately rely on the AI over time, not on whether their capacity for independent reasoning is maintained or reduced through repeated interaction"

Discussion¶

Related Claims¶

• Error timing and stakes shape trust in AI systems: later errors may be more damaging, early errors may matter more, and trust recovers quickly after good post-error performance, with high-stakes errors causing abrupt collapse — related
• The review reports some evidence that over-dependence on AI contributes to cognitive offloading and decreased independent reasoning, while guided AI use can be helpful — related
• A systematic review reports that over-reliance on AI dialogue systems fosters complacency and dependency and may weaken students' decision-making — related
• AI support can improve task performance while weakening users' metacognitive accuracy and leaving their own reasoning unimproved — related
• AI systems relying on black-box models should not be used for high-stakes applications — related