Home/Events/AI-Generated Scientific Paper Experiment Using ChatGPT and Gemini

AI-Generated Scientific Paper Experiment Using ChatGPT and Gemini

Emerging
Confidence
80%
Impact: 70%
Updated 6m ago

Consensus Brief

Benjamin Skuse conducted an experiment to determine if AI-generated research could pass as legitimate academic work. The findings indicate a significant rise in AI-generated content in academic publishing, with predictions that 57% of published articles will show LLM influence by 2025. The experiment revealed the ease of creating convincing fake papers using off-the-shelf AI tools.

Sourced from
Primary: IEEE Spectrum

What Changed Since Last Update

6m ago

New corroborating source added: IEEE Spectrum published an update on Tue, 27 Oc ("I Used AI to Fake a Scientific Paper. Here’s What Happened").

Claim Ledger

4 claims tracked across sources

Confirmed Fact

By 2025, an estimated 57 percent of all published academic articles exhibited detectable evidence of large language model (LLM) influence.

Confirmed Fact

Over 12 percent of Nature Communications reviews were found to be completely AI-generated in 2025.

Independent Finding

Hundreds of thousands of fake, often AI-generated, publications are now produced annually.

Confirmed Fact

For every 40 papers showing statistical evidence of AI usage in 2025, only one formally disclosed it.

Role-Based Impact Analysis