Poorly grounded GenAI self-study risks fabricated content and inferior learning relative to curated resources¶

Claim · All claims
Evidence · 1 study (2 entries) · review r3 · q2
Currency · newest study 2026 · checked against its sources 2026-10-20 (automated, 2 of 2 entries) · not reviewed by a person

Subclaims¶

q2 i? During an initial ChatGPT-3.5 phase only 6 of 23 students correctly learned the Randleman criteria, while all 23 answered correctly in a subsequent internet-search phase, and the model had fabricated a similarly named syndrome. → Zhao J 2026 q2 i? Blended urology teaching outperformed ChatGPT-only learning, and in other comparisons textbooks or video instruction performed as well or better. → Zhao J 2026 (2)

Evidence¶

Zhao J 2026¶

Zhao J, Yang K, Gao H and Jiang G (2026). Generative artificial intelligence and the transformation of medical education: a scoping review. Front. Educ. 11:1907589. https://doi.org/10.3389/feduc.2026.1907589

q2 · i? · review · r3

The review reports this behavioral comparison from Tuttle et al. as a stark example of hallucination becoming learned misinformation during ungrounded self-study.

"Only 6 of 23 students correctly learned the Randleman criteria during an initial ChatGPT-3.5 phase; all 23 answered correctly in a subsequent internet-search phase, and the model had fabricated a similarly named syndrome (Tuttle et al., 2024)."

Zhao J 2026 (2)¶

Zhao J, Yang K, Gao H and Jiang G (2026). Generative artificial intelligence and the transformation of medical education: a scoping review. Front. Educ. 11:1907589. https://doi.org/10.3389/feduc.2026.1907589

q2 · i? · review · r3

The review reports this comparison among negative and null results showing the risk of poorly grounded self-study; other comparisons found ChatGPT faster without knowledge improvement.

"Blended urology teaching outperformed ChatGPT-only learning (Digiacomo et al., 2025)."

Discussion¶

Related Claims¶

• The most frequently cited ethical concerns about GenAI in research were hallucinated facts, fabricated references, and misleading content — related
• Students report negative experiences with ChatGPT including inaccurate outputs, lack of emotional connection, and risk of overreliance, which constrain competence — related
• AI output errors are frequent: 44.5% fabricated references in drafted research proposals and 5–13% unsafe chatbot medical answers — related
• Students raised ethical concerns about GenAI-assisted data comic creation, most frequently misinformation, bias, ownership, and reduced trustworthiness — related
• Students' main AI challenges are content reliability and depth, and most expect AI to serve as tutor, creative partner, and critical thinking partner rather than mere content generator — related
• Les citations bibliographiques produites par ChatGPT sont inexistantes ou incorrectes dans 23 % des cas, compromettant le circuit de validation scientifique — related
• The most frequent limitation of AI in instructional design is content reliability and transparency (16 of 28 studies, 57.14%) — related
• Gen-AI hallucinates sources and values and produces convincing but incorrect explanations, risking student misinformation — possibly the same claim (merge candidate)
• Source-anchored instruction pairs reduce hallucination: they achieve mean Teacher Fidelity 1.4 points higher than non-anchored pairs, and models trained on anchored subsets produce 23% fewer factual hallucinations — related
• Review-attributed evidence: fabricated model-generated references clustered in disciplinary journals in a student verification exercise — related
• AI-assisted scientific writing assignments risk fabricated references and altered chemical meaning — related
• Purpose-built chatbot effects on educational performance are mixed, with some studies showing no benefits or worse outcomes than comparison conditions — related
• Undergraduate students’ genAI knowledge contains specific confident misconceptions, most prevalently that AI chatbots perform web searches — related
• Of 178 reference sources produced by AI dialogue systems, 69 lacked a DOI and 28 could not be found via Google search and had no existing DOI — related
• Les LLMs utilisés comme juges d'évaluation scientifique présentent douze types de biais qui désavantagent les productions s'écartant des canons dominants — related
• ChatGPT could not execute R code for statistics MCQs, achieving 60% accuracy when its code was run with corrected dataset naming — related