ACM WebConf 2026 ยท Dubai

Can humans tell AI news from real journalism?

A two-part research platform: RogueGPT generates controlled AI news stimuli. JudgeGPT measures whether humans can detect them. The answer might surprise you.

JudgeGPT research platform
๐Ÿ”ฌ Try the Live Demo ๐Ÿ“„ Read the Papers GitHub
3,278
News Fragments
37
Model Configs
10
AI Providers
4
Languages
3
Papers

Research pipeline

1
๐Ÿงช

Generate

RogueGPT

Controlled AI news generation across 10 models, 4 languages, 5 journalistic styles.

โ†’
2
๐Ÿ—„๏ธ

Store

MongoDB Atlas

Full provenance metadata persisted with every fragment. Reproducible filtering by any variable.

โ†’
3
โš–๏ธ

Evaluate

JudgeGPT

Humans judge authenticity in a controlled experiment. Confidence, reasoning, and demographics captured.

โ†’
4
๐Ÿ“Š

Analyze

Analysis Pipeline

Statistical analysis of human susceptibility across models, languages, and demographics.

Published research

Industrialized Deception paper
WWW '26 Companion ยท Dubai

Industrialized Deception: The Collateral Effects of LLM-Generated Misinformation

How LLM-generated misinformation disrupts digital platforms, trust networks, and information ecosystems at scale.

Eroding Truth Default paper
WWW '26 Companion ยท Dubai

Eroding the Truth-Default: Human Susceptibility to Foundation Model Hallucinations

Causal analysis of how humans perceive AI-generated news. Can they tell the difference? Spoiler: barely.

Blessing or Curse survey
Foundation Survey

Blessing or Curse? A Survey on the Impact of Generative AI on Fake News

The foundational literature survey that motivated this research program.

Two repos, one mission

โš–๏ธ
JudgeGPT
โญ 8

Human evaluation platform. Streamlit-based experiment collecting authenticity judgments, confidence levels, and reasoning.

PythonStreamlitMongoDBReact
๐Ÿงช
RogueGPT
โญ 5

Controlled stimulus generator. Produces AI news fragments with explicit provenance across 10 models and 6 providers.

PythonOpenAIMongoDBMulti-LLM

Help us measure human perception

Take the survey. Can you tell which news was written by AI? Most people can't.

๐Ÿ”ฌ Take the Survey Star on GitHub โญ ๐Ÿ“ฆ Zenodo Dataset