MétaCan
Menu
Back to cohort
Record W4415300695 · doi:10.1038/s41746-025-02008-z

When helpfulness backfires: LLMs and the risk of false medical information due to sycophantic behavior

2025· article· en· W4415300695 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

Venuenpj Digital Medicine · 2025
Typearticle
Languageen
FieldComputer Science
TopicTopic Modeling
Canadian institutionsArtificial Intelligence in Medicine (Canada)
FundersGoogle ResearchNational Center for Advancing Translational SciencesEuropean CommissionHarvard CatalystNational Cancer InstituteNational Institutes of HealthHarvard UniversityPatient-Centered Outcomes Research Institute
KeywordsHelpfulnessVulnerability (computing)Consistency (knowledge bases)SuspectRisk assessmentBenchmark (surveying)Baseline (sea)Sophistication

Abstract

fetched live from OpenAlex

Large language models (LLMs) exhibit a vulnerability arising from being trained to be helpful: a tendency to comply with illogical requests that would generate false information, even when they have the knowledge to identify the request as illogical. This study investigated this vulnerability in the medical domain, evaluating five frontier LLMs using prompts that misrepresent equivalent drug relationships. We tested baseline sycophancy, the impact of prompts allowing rejection and emphasizing factual recall, and the effects of fine-tuning on a dataset of illogical requests, including out-of-distribution generalization. Results showed high initial compliance (up to 100%) across all models, prioritizing helpfulness over logical consistency. Prompt engineering and fine-tuning improved performance, improving rejection rates on illogical requests while maintaining general benchmark performance. This demonstrates that prioritizing logical consistency through targeted training and prompting is crucial for mitigating the risk of generating false medical information and ensuring the safe deployment of LLMs in healthcare.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.002
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Other design · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.922
Threshold uncertainty score0.250

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.002
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0000.001
Open science0.0010.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.008
GPT teacher head0.247
Teacher spread0.239 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it