How AI perpetuates misinformation through “digital fossils” in scientific literature

Artificial intelligence systems are increasingly spreading and propagating errors through our collective knowledge base, creating “digital fossils” that become permanently embedded in scientific literature. The case of “vegetative electron microscopy” – a nonsensical term born from scanning and translation errors that has appeared in 22 scientific papers – reveals how AI systems can amplify and perpetuate misinformation. This phenomenon highlights a growing concern about the integrity of our digital knowledge repositories and the reliability of AI-generated content in scientific contexts.

The big picture: “Vegetative electron microscopy” emerged through a remarkable coincidence of unrelated errors in document digitization and translation, revealing how technical-sounding but meaningless terms can infiltrate scientific literature.

• The term originated when two papers from the 1950s were improperly digitized, erroneously combining “vegetative” from one column with “electron” from another column of text.
• Decades later, the nonsensical term appeared in Iranian scientific papers due to a translation error, as the Farsi words for “vegetative” and “scanning” differ by only a single dot.

Key details: The problematic term has spread to nearly two dozen scientific papers, triggering retractions and corrections while highlighting deeper issues with information integrity.

• Google Scholar currently shows “vegetative electron microscopy” appearing in 22 scientific papers, including one that faced a contested retraction from Springer Nature and another that received a correction from Elsevier.
• The term has also appeared in news articles discussing subsequent scientific integrity investigations, further cementing its presence in the digital knowledge ecosystem.

Why this matters: These “digital fossils” demonstrate how AI systems can inadvertently preserve and amplify errors throughout our information landscape, creating long-term challenges for scientific accuracy.

• Like biological fossils trapped in rock, these digital artifacts can become permanent fixtures in our knowledge repositories, nearly impossible to completely remove once they’ve been integrated into training datasets.
• This case offers a troubling glimpse into how AI systems perpetuate and amplify errors, potentially undermining scientific progress and the reliability of digital knowledge.

Implications: The persistence of such errors raises questions about the long-term integrity of scientific literature and the need for better safeguards in AI development.

• As AI systems increasingly incorporate scientific literature into their training data, the risk of these errors being perpetuated and amplified grows significantly.
• The scientific community now faces challenges in developing methods to identify and correct such digital fossils before they become permanently embedded in our collective knowledge.