Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 10, 2025Journal of CommunicationOpen Access

The (in)efficacy of AI personas in deception detection experiments

View Full Paper
Ask AI
Bookmark
Share

Authors

DMDavid M. MarkowitzTLTimothy R. Levine

Discussion

Loading...

Member takes

Overview

Experiments show AI personas struggle with deception detection in varied communication modalities, highlighting limitations.

Key Points

  • AI demonstrated a maximum accuracy of 57.7% in detecting truths and lies about friends.
  • Notably, accuracy dropped to 15.9% for human interviews at rates above 70%, indicating significant limitations.
  • The study used 12 experiments across different communication types and conditions to evaluate AI performance.
  • Results suggest caution in using large language models for lie detection due to discrepancies with human study findings.

Cite This Study

Markowitz et al. (2025) studied this question.

synapsesocial.com/papers/68c194029b7b07f3a06188e1https://doi.org/10.1093/joc/jqaf034
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Caution with AI deception: AI could serve as the scapegoat of Human!2024
  2. 2Deciphering Deception: How Different Rhetoric of AI Language Impacts Users’ Sense of Truth in LLMs2024 · 13 citations
  3. 3It’s not what you say, but how you say it: Using Artifact Intelligence to Detect Deception in Natural Speech in an HR Recruiting Context2026
  4. 4Misinformation Detection by AI Chatbots and Humans2025
  5. 5Lying toward AI agent: the roles of type of lie, moral disengagement and perceived emotional ability2025