Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
March 5, 2026Open Access

Unmasking the Algorithm: A Comprehensive Meta-Analysis and Predictive Modeling of Human versus Machine Efficacy in AI Text Detection

View Full Paper
Ask AI
Bookmark
Share

Authors

OTOwen R. Thornton

Discussion

Loading...

Member takes

Overview

Meta-analysis evaluates text detection efficacy in AI versus human authorship, suggesting a new predictive modeling framework.

Key Points

  • The aim is to assess text detection performance between human evaluators and AI systems, while developing a predictive modeling framework.
  • Conducted a meta-analysis of peer-reviewed studies on text detection efficacy.
  • Aggregated performance metrics from human evaluators and commercial AI detectors.
  • Developed a predictive modeling framework using supervised and unsupervised learning techniques.
  • Established a unified overview of detection performance across different evaluation methods.
  • Identified significant vulnerabilities in current detection systems, especially against adversarial tactics.
  • Found human evaluators frequently misattribute synthetic texts to human authorship.

Cite This Study

Owen R. Thornton (2026) studied this question.

synapsesocial.com/papers/69a91e1fd6127c7a504c1b87https://doi.org/10.17615/qep9-mm62
View Full Paper
Ask AI
Bookmark
Share