AI detectors are poor western blot classifiers: a study of accuracy and predictive values

The recent rise of generative artificial intelligence (AI) capable of creating scientific images presents a challenge in the fight against academic fraud. This study evaluates the efficacy of three free web-based AI detectors in identifying AI-generated images of western blots, which is a very commo...

Full description

Saved in:
Bibliographic Details
Main Author: Romain-Daniel Gosselin
Format: Article
Language:English
Published: PeerJ Inc. 2025-02-01
Series:PeerJ
Subjects:
Online Access:https://peerj.com/articles/18988.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850191335028424704
author Romain-Daniel Gosselin
author_facet Romain-Daniel Gosselin
author_sort Romain-Daniel Gosselin
collection DOAJ
description The recent rise of generative artificial intelligence (AI) capable of creating scientific images presents a challenge in the fight against academic fraud. This study evaluates the efficacy of three free web-based AI detectors in identifying AI-generated images of western blots, which is a very common technique in biology. We tested these detectors on AI-generated western blot images (n = 48, created using ChatGPT 4) and on authentic western blots (n = 48, from articles published before the rise of generative AI). Each detector returned a very different sensitivity (Is It AI?: 0.9583; Hive Moderation: 0.1875; and Illuminarty: 0.7083) and specificity (Is It AI?: 0.5417; Hive Moderation: 0.8750; and Illuminarty: 0.4167), and the predicted positive predictive value (PPV) for each was low. This suggests significant challenges in confidently determining image authenticity based solely on the current free AI detectors. Reducing the size of western blots reduced the sensitivity, increased the specificity, and did not markedly affect the accuracy of the three detectors, and only slightly improved the PPV of one detector (Is It AI?). These findings highlight the risks of relying on generic, freely available detectors that lack sufficient reliability, and demonstrate the urgent need for more robust detectors that are specifically trained on scientific contents such as western blot images.
format Article
id doaj-art-4f37982e5b7b4334b9d4ec07f26a1f53
institution OA Journals
issn 2167-8359
language English
publishDate 2025-02-01
publisher PeerJ Inc.
record_format Article
series PeerJ
spelling doaj-art-4f37982e5b7b4334b9d4ec07f26a1f532025-08-20T02:14:57ZengPeerJ Inc.PeerJ2167-83592025-02-0113e1898810.7717/peerj.18988AI detectors are poor western blot classifiers: a study of accuracy and predictive valuesRomain-Daniel GosselinThe recent rise of generative artificial intelligence (AI) capable of creating scientific images presents a challenge in the fight against academic fraud. This study evaluates the efficacy of three free web-based AI detectors in identifying AI-generated images of western blots, which is a very common technique in biology. We tested these detectors on AI-generated western blot images (n = 48, created using ChatGPT 4) and on authentic western blots (n = 48, from articles published before the rise of generative AI). Each detector returned a very different sensitivity (Is It AI?: 0.9583; Hive Moderation: 0.1875; and Illuminarty: 0.7083) and specificity (Is It AI?: 0.5417; Hive Moderation: 0.8750; and Illuminarty: 0.4167), and the predicted positive predictive value (PPV) for each was low. This suggests significant challenges in confidently determining image authenticity based solely on the current free AI detectors. Reducing the size of western blots reduced the sensitivity, increased the specificity, and did not markedly affect the accuracy of the three detectors, and only slightly improved the PPV of one detector (Is It AI?). These findings highlight the risks of relying on generic, freely available detectors that lack sufficient reliability, and demonstrate the urgent need for more robust detectors that are specifically trained on scientific contents such as western blot images.https://peerj.com/articles/18988.pdfResearch integrityAI detectionPaper millsFraudResearch ethicsAccuracy study
spellingShingle Romain-Daniel Gosselin
AI detectors are poor western blot classifiers: a study of accuracy and predictive values
PeerJ
Research integrity
AI detection
Paper mills
Fraud
Research ethics
Accuracy study
title AI detectors are poor western blot classifiers: a study of accuracy and predictive values
title_full AI detectors are poor western blot classifiers: a study of accuracy and predictive values
title_fullStr AI detectors are poor western blot classifiers: a study of accuracy and predictive values
title_full_unstemmed AI detectors are poor western blot classifiers: a study of accuracy and predictive values
title_short AI detectors are poor western blot classifiers: a study of accuracy and predictive values
title_sort ai detectors are poor western blot classifiers a study of accuracy and predictive values
topic Research integrity
AI detection
Paper mills
Fraud
Research ethics
Accuracy study
url https://peerj.com/articles/18988.pdf
work_keys_str_mv AT romaindanielgosselin aidetectorsarepoorwesternblotclassifiersastudyofaccuracyandpredictivevalues