Skip to content
Submit incident
AII-1001-90203
ReportedReported by fewer than three independent publishers, with no official finding.

Springer Nature book on machine learning published with fabricated citations

Springer Nature published a machine learning textbook containing numerous fabricated or incorrect citations, with two-thirds of sampled citations either nonexistent or substantially erroneous. Three researchers confirmed works attributed to them were fake or misrepresented, with patterns consistent with large language model-generated content.

18 April 2025 · Govindakumar Madhavan · Reliability and hallucination
18 April 2025
AII-1001-90203View source ↗
Quick facts
Category
Reliability and hallucination
Source
retractionwatch.com
Date
18 April 2025
Drafted by
Raidu Incident Detection

What happened

Springer Nature published 'Mastering Machine Learning: From Basics to Advanced' by Govindakumar Madhavan in April 2025. Retraction Watch checked 18 of the 46 citations in the 257-page book and found two-thirds either did not exist or contained substantial errors.

Three researchers contacted confirmed works attributed to them were fake or contained substantial errors. Yehuda Dar stated a cited paper was an arXiv preprint incorrectly listed as published in IEEE Signal Processing Magazine. Aaron Courville noted a cited section 'doesn't seem to exist' at the pages specified. Dimitris Kalles confirmed he did not author a work listed under his name.

Reported impact

The book was priced at $169 for an introductory ebook on machine learning, potentially misleading readers about its citations and reliability.

Index analysis

Fabricated and error-prone citations are a known hallmark of text generated by large language models like ChatGPT. LLMs generate content based on training data rather than searching literature databases, potentially producing citations that appear legitimate but contain fabricated content.

Madhavan did not directly answer whether an LLM was used to generate the book's text, stating instead that 'reliably determining whether content is AI generated remains a challenge.'

Governance controls

Springer Nature provides policies and guidance about the use of AI to its authors.

Classification

Category
Reliability and hallucination
Industry
Education and training
Harm status
Actual harm
Intent
Unknown
Timing
Post-deployment
Harm types
Reputational
Harmed
Consumers, Business
Failure modes
Fabricated citation, Hallucination
Consequences
None recorded
AI systems
ChatGPT

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

retractionwatch.com
Springer Nature book on machine learning published with fabricated citations