This letter argues for the use of confabulation over hallucination to better shape clinical trust in AI errors, emphasizing practical clinical implications.
This Letter responds to the ongoing debate on the terminology for large language model errors by arguing that “hallucination” mislocates the error in the perceptual domain and that “confabulation” is a better-calibrated conceptual borrowing. It further proposes a functional typology distinguishing ordinary from delusional confabulation, with practical implications for clinical settings.
No takes yet. Share an insight, caveat, or question.
Alfredo Vannacci (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: