A Paper I Never Wrote Showed Up in a Journal
One day, Alan Said, a professor of computer science at the University of Gothenburg in Sweden, got an automatic Google Scholar alert email. It said a new paper of his had been published in a journal.
He scratched his head, opened the link, and found a paper called "Intelligence Systems Supporting Sustainable Global Development" listed there, apparently.
The topic itself looked natural. Said has long researched AI, machine learning and the sustainability of recommender systems. But he'd never written or submitted any such paper.
Once he skimmed the body text and references, he realized right away it was something an AI had made up, they say. The sentences looked polished on the surface, but there was no verified data behind them, and the citation list was sloppy.
In an interview with Times Higher Education, he described the piece as mass-produced "slop" with no real content.
✦ Same Paper, Only the Author Changed
Three versions in three days
In the end, Professor Said reported it to the domain registrar of AIJRI (American International Journal of Research and Innovation), the journal that listed the paper, and the reply came back that the paper had been deleted.
But when he went to the site, the paper was still there and the only thing that had changed was the author name. In the screenshots Said made public, the author name on this paper changed again two days later, to yet another person.
Within three days, the same AI paper had been listed in three versions with only the author swapped out.
By the way, the March-April 2026 issue PDF (38 pages) of AIJRI now lists the paper's author as Dr. Vaelira Norellin, with "Sweden" as the nationality. The affiliation field says "Virelmont Institute of Computational Sustainability, Melvaren, Asterovia."
Top Scholars Whose Names Ghost Journals Stole
✦ Four Pieces of Digital Forensic Evidence
Cloned journals AIJRI, AJAMRI, AJFMR
When Said chased down the traces, AIJRI turned out to be part of a huge cloning network.
It was tangled up with two sister journals, "AJAMRI (American Journal of Advanced Multidisciplinary Research and Innovation)" and "AJFMR (American Journal For Multidisciplinary Research)," whose page designs and even editorial board lists were near carbon copies.
Their server addresses were bundled in the same IP range at the same hosting company, and all the domains were newly registered in 2026. The sites were stacked with back issues whose dates were pushed back to look like they were published in 2019 and 2023.
The journal ID number (E-ISSN) was still the temporary blank "XXXX-XXXX," too.
These journals list the same editor-in-chief name, "Denish Brown," and the same US address. But that address isn't a publisher's headquarters. It's a virtual office address that rents out just a mailing address by the month.
Their payment instructions lead with transfers to an Indian bank account and an Indian payment processor.
Behind how a site registered in 2026 can pose as a journal that's been publishing since 2019 and 2023 is the automatic collection structure of Google Scholar. Google Scholar isn't a place that reviews or certifies journals.
If you match the publication date meta tag and the author and abstract format on a public page, the crawler simply collects it automatically.
When a ghost journal writes a past year in its HTML tags and attaches a renowned scholar's name, the search robot pushes it into profiles and alert systems as if it were an academic record written by the real scholar.
✦ Publication Fee Comparison
AIJRI and AJFMR, for non-Indian authors
In the end, the target is probably early-career researchers' wallets.
AIJRI and AJFMR charge submitters a fee per paper, and AJFMR advertises that it publishes the paper within a day or two of the payment receipt being uploaded, and even issues a digital certificate.
Said worried that a doctoral student might see a famous professor's name and mistake these ghost journals for legitimate ones. Their precious research and publication fees would go down the drain.
✦ Scholars Whose Names Were Stolen
In fact, these ghost journals have been building authority by stealing the names of leading AI scholars.
The January-February 2019 issue page of AJAMRI has a fintech paper under the name of Stuart Russell of UC Berkeley and a smart agriculture paper under the name of Jure Leskovec of Stanford, and the November-December 2023 issue carries a quantum and green computing paper under the name of Fei-Fei Li.
The paper has an odd AI-style sentence that strings three concepts together with plus signs. They say the names of Christopher Manning and Michael I. Jordan were used on dozens of papers, too.
Professor Leskovec told reporters flat out, "I never wrote or submitted that paper, and I never gave permission to use my name."
Even Kanakaraju Manupati of Virginia Commonwealth University, whose name is listed as an "editorial board member" of the ghost journals, said, "I didn't even know the journal existed, and I never agreed to be appointed."
When the media asked for their position, the three journals never replied up to the time of reporting.
Cases Where Ghost Journals Were Punished
✦ Investigation vs. Court Rulings
The digital evidence is clear: contradictions in domain registration dates, fake ID numbers, virtual offices, Indian bank payment routes, victim testimony and more.
But maybe because they sit behind overseas servers and virtual bases, so far no police investigation results or criminal charges against the AIJRI, AJAMRI and AJFMR operators have been confirmed.
Still, there's a precedent where a similar scheme was judged in court. The Federal Trade Commission (FTC) sued OMICS Group, a predatory publisher based in India, on charges of using scholars' names without consent to collect publication fees.
In March 2019, a federal district court in Nevada found that OMICS's journal and conference operations involved "deceptive acts or practices" under the FTC Act.
It then ordered about $50.13 million in monetary relief, matching the estimated consumer losses, along with a permanent injunction against the illegal conduct.
✦ GIJIR Journal Verification Results
The 53 papers with the fewest in-text citations, Spinellis 2025
Academia is stepping up to verify, too.
After finding a fake AI paper written under his own name, Professor Diomidis Spinellis of the Athens University of Economics and Business in Greece crawled the entire journal "GIJIR" and published a case study in May 2025.
Of the 53 papers with the fewest in-text citations, at least 48 were assessed as probably mostly AI-generated, and they say even the names of scholars who had already died were listed as authors.
In April 2026, Dr. Eric Topol, a cardiologist in the United States, found a fake paper and exposed it, and within hours the entire site of the journal and its publisher had vanished.
For early-career researchers who are just starting to get their names out there, it must be really tempting: pay a cheap listing fee and you're published in a journal alongside famous scholars.
On top of that, with no tough review and a listing in just a day or two, it's easy to see the appeal. They probably spent a long, long time on research to write that one paper.
How crushing it must feel to find out later that the place you published is a ghost journal.
Right now, AI-written papers are sloppy enough that they give themselves away fast, but down the road, once AI gets as good at writing papers as humans are, I worry ghost journals will run even wilder.
That's all for this post. See you next time.
Related posts
- A Fake £154 Million Takeover Bid, Made With ChatGPT
- 200-Page AI Complaints Are Burying UK Council Officials
- AI-Faked Photo Won a DoorDash Pizza Refund. CCTV Caught It
🌐 Korean version: "예? 저는 그런 논문을 쓴 적이 없는데요?" AI 가짜논문으로 도배되는 학술지의 실체
🌐 Japanese version: 「そんな論文は書いていません」AI偽論文だらけの幽霊学術誌
No comments:
No comments yet. Want to leave the first one?
Post a Comment