He Never Wrote It.
It Ranked First Anyway.
In December 2023, a predatory journal published a fabricated article under a real professor’s byline: his name, his university, a working DOI (Digital Object Identifier). He never wrote it, submitted it, or knew it existed until a colleague turned it up. His investigation later found other established academics named in the journal’s archive, including two who were already dead when articles appeared under their names — real, checkable identities attached to work the evidence indicates they did not submit. The chain that registered, indexed, and ranked
this article carried Spinellis’s name without establishing
whether he had authorized the byline.
In December 2023, a journal calling itself the Global International Journal of Innovative Research published a comparative-strategy article on multinational corporations under a real name: Diomidis Spinellis, a professor at TU Delft’s Department of Software Technology, also affiliated with the Athens University of Economics and Business. It carried his correct affiliation, pages 240 through 246, and a working DOI. He did not write it. He did not submit it. He did not know it existed until a colleague, wanting a summary for an MBA newsletter, found it through a Google Scholar search and asked him for one.
Exposure did not erase the record cleanly. After Spinellis discussed the article publicly and notified other named authors, the journal temporarily disappeared. It later returned without the article attributed to him. The DOI remained registered, and archived copies survived. Removal corrected one surface. It did not recall every copy of the lie.
Spinellis did what most people whose names get stolen this way never get the chance to do: he investigated it as a scientist. His peer-reviewed forensic study, published in Research Integrity and Peer Review in 2025, crawled the journal’s entire website and extracted its article metadata, citation patterns, affiliations, and contact emails. He then manually examined two selected sets of papers and submitted a smaller, systematically selected subset to Turnitin for an AI-generation score. Of the fifty-three manually examined articles with the fewest probable in-text citations, he judged at least forty-eight to be probably mostly AI-generated — Turnitin’s own scores on the tested subset ran high. That finding stands on its own: a fact about the journal, not about him.
The narrower, stranger pattern sits inside the journal’s earliest volumes: apparently fabricated articles attributed to real, tenured academics at real institutions — Washington University, Texas A&M, UC Berkeley, USC’s Keck School of Medicine, and Spinellis’s own university among them. Two of the borrowed names belonged to scholars who were already dead when the articles under their names went to press: Christodoulos A. Floudas and Georges Guiochon. Spinellis’s own paper states the obvious gap plainly — every victim in this category was already a tenured, established researcher, so “there is also a lack of a motive to publish an article containing blatantly fake scientific research.” The fabrication wasn’t self-inflicted. It was manufactured, and then worn.
None of it required breaching a technical system. The journal carries a valid ISSN. Its publisher, identified through Spinellis’s own Crossref queries as an operation called Yayasan Banu Samsudin running eleven journal titles from the same registration, deposited article metadata through its Crossref membership, producing working DOIs, and Google Scholar indexed material from the journal alongside genuine research. Crossref’s own membership terms require depositing accurate metadata and registering only content the member has the legal rights to register. This record entered the scholarly infrastructure under Spinellis’s name anyway. In the documented path from publisher metadata to DOI registration to search index, no author-consent check stopped it.
What the record can’t tell apart
Spinellis’s established record gave the stolen byline immediate credibility. Decades of real, indexed, citable work were already attached to the name. The other borrowed names produced the same effect: they were real, established, and already carried weight before anyone forged anything.
The observed pipeline performed its ordinary functions, and those functions did not stop the false attribution. The publisher deposited metadata. Crossref registered the DOI. Google Scholar indexed the result. At every documented step, the record moved forward without Spinellis’s consent being established.
Once this byline cleared registration, the systems downstream carried it without resolving where it came from. Filed as Authority Misclassification, one of the five clinical characteristics of Digital Derangement Syndrome™: ranking and citation were built to move a credited name faster. Here, they moved a stolen one just as fast. Ranking and citation stop being a defense. They become the delivery mechanism.
What a dense record doesn’t fix
Spinellis caught this one because he happened to look — a colleague’s request, a Google Scholar link, and enough of his own established record to know instantly that the writing wasn’t his. His study, published in May 2025, documents that the fabricated article appeared first in Google Scholar for the Google Scholar query “Spinellis business.” A fabrication with his name on it entered the record doing whatever work a citation does for whoever encountered it next — and nothing in the documented chain stopped it before it entered the record.
Fifty-one editions into this record, the pattern this one belongs to is the sharpest version yet of something every prior edition has circled: recognition infrastructure and the record it is supposed to represent are not the same system, and this case shows that the rules did not force them to agree in practice. Spinellis had the standing, and the scientific training, to investigate his own theft and publish the proof. A professional can remain unaware indefinitely, because discovery still depends on someone thinking to look.
Spinellis, D., “False authorship: an explorative case study around an AI-generated article published under my name,” Research Integrity and Peer Review 10, 8 (2025) — peer-reviewed, open access. Direct quotations verified against the full text: the December 2023 publication of “Global Business Strategies in the Digital Age” under Spinellis’s name (GIJIR, pp. 240–246, DOI 10.59613/global.v1i3.42); the finding that “at least 48” of 53 sampled articles “appeared to be AI-generated”; the misattribution to Christodoulos A. Floudas and Georges Guiochon, “authors who had died several years before the articles under their name were published”; the publisher identification as “Yayasan Banu Samsudin”; and the Google Scholar ranking as documented at the time of writing — doi.org/10.1186/s41073-025-00165-z.
TU Delft, “Scientific Study Exposes Publication Fraud Involving Widespread Use of AI” — institutional press release corroborating the study’s findings — tudelft.nl/en/2025/eemcs/scientific-study-exposes-publication-fraud-involving-widespread-use-of-ai.
Crossref, Member Obligations — members must “deposit accurate metadata,” and, per clause 3(c) of Crossref’s terms, “will not deposit or register Metadata for any Content for which the Member does not have legal rights to do so” — crossref.org/membership/.