Dialogue-derived interactional cues robustly encode ground-truth identity, achieving near-ceiling AI identification under cross-validation¶
Claim · All claims
Evidence · 1 study · causalr2·q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q3 i? A Top-1 identification analysis correctly identified the AI teammate in 95.3% of triads, far exceeding the 33.3% chance baseline, and performance remained high when timing features were excluded. → Lixiang Yan 2026
Evidence¶
Lixiang Yan 2026¶
Lixiang Yan, Yueqiao Jin, Xibin Han, Dragan Gasevic. (2026). Socially fluent AI decouples conversational signals from source identity in online interaction. https://arxiv.org/abs/2605.23426 doi:10.48550/arxiv.2605.23426
q3 · i? · causal · r2
A Top-1 identification analysis per triad "correctly identified the AI teammate in 95.3% of triads" (95% CI [0.91, 1.00]) versus a 33.3% chance baseline; performance remained comparably high without timing features (96.6%). Permutation tests confirmed the observed AUC lay far in the upper tail (p < .001).
"At the group level, a Top-1 identification analysis, selecting the single teammate with the highest predicted probability per triad, correctly identified the AI teammate in 95.3% of triads (95% CI [0.91, 1.00]), far exceeding the chance baseline of 33.3%."
Discussion¶
Related Claims¶
- Participants fail to distinguish undisclosed AI teammates from human teammates above chance in synchronous text-based group interaction — reports the opposite
- Judgment-truth dissociation: the same cues that predict ground-truth identity are only weakly incorporated into explicit identity judgments — related
- Systems adapt to the specified learner persona in ways human raters can perceive, scoring well above chance in persona identification — related
- Representational similarity analysis shows identity judgments are organised around subjective impressions, orthogonal to diagnostic cue structure — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
cues-encode-identity-near-ceiling-classification |
title |
Dialogue-derived interactional cues robustly encode ground-truth identity, achieving near-ceiling AI identification under cross-validation |
description |
Dialogue-derived interactional cues robustly encode ground-truth identity, achieving near-ceiling AI identification under cross-validation |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2026 |
recheck_by |
2029 |
generated |
by process:wiki-ingest on 2026-10-11 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — lixiang-yan-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.