AI-generated junk papers falsely use the names of scholars such as Hinton and Fei-Fei Li, and pay-to-publish journals use these names to attract submissions
A group of newly registered pay-to-publish journals has been exposed for mass-publishing AI-generated junk papers under the names of well-known scholars such as Geoffrey Hinton, Yann LeCun, Fei-Fei Li, and Andrew Ng, all without the knowledge of the named individuals. After Alan Said, a professor at the University of Gothenburg, discovered the impersonation papers, the paper's author was changed twice within two days, and the editor-in-chief and editorial board lists of the same batch of journals were also alleged to have been misappropriated. Christopher Manning called on Google Scholar not to index these papers, and the report also noted that such content may flow back into model training corpora.
Why read this
AI score 66
The report outlines how impersonation journals use AI to mass-generate papers and attach the names of well-known scholars, as well as why the indexing stage becomes a risk entry point.
Topics
Body · AI translation
One day, Alan Said, a professor of computer science at the University of Gothenburg, received a Google Scholar alert: a new paper of his had been published.
The paper was titled "Intelligence Systems Supporting Sustainable Global Development" and appeared in a journal called the American International Journal of Research and Innovation (AIJRI). The title was tangentially related to his research area of "AI and sustainability," and the byline was indeed his.
The problem: he didn't write the paper.
"It's not strange for me to write a paper in this area, but one look at the paper itself and its references is enough to tell it was AI-generated. It has no content, it's just a pile of slop," Said told Times Higher Education (THE).
Following this paper, THE uncovered a much larger operation: a group of obscure journals is mass-publishing AI-generated junk papers under the names of professors at top universities such as Stanford and Berkeley. Those whose names were "attached" include many famous AI researchers, among them Geoffrey Hinton, Yann LeCun, Yoshua Bengio, Fei-Fei Li, Christopher Manning, Michael I. Jordan, Andrew Ng and Stuart Russell.
The author was removed, but the paper stayed, and the name was swapped and hung up again
But the paper did not disappear; the author was simply changed from Said to someone else. According to screenshots Said provided to THE, two days later the author was changed again. In this way, the same paper was successively attached to three different people's names.
This detail is quite telling: for journals like these, the content of the paper does not matter at all; what matters is the name attached to it. Names can be swapped at any time, like changing a model photo on a product page.
Said then found two more "sibling journals": the American Journal of Advanced Multidisciplinary Research and Innovation (AJAMRI) and the American Journal For Multidisciplinary Research (AJFMR). The three journals share a similar visual style and the same editor-in-chief and editorial board members.
Among them, AJAMRI has already published dozens of papers attributed to well-known scholars, a lineup that reads like an AI "all-star team": Fei-Fei Li and Christopher Manning of Stanford, and Michael I. Jordan and Stuart Russell of Berkeley.
Jure Leskovec, a professor of computer science at Stanford University, is also on the list. He told THE unequivocally: "I can confirm that I did not write this paper, did not participate in its submission, and did not authorize the use of my name. I was not even aware of its existence until I received your email. This is quite concerning."
A close look at these three journals reveals that the flaws are not hard to find.
All three websites' domains were only registered in 2026, yet the sites display archives of papers "published several years ago." They all describe themselves as "peer-reviewed, open-access, multidisciplinary" journals and openly charge publication fees.
The fee rules are also worth a look. AIJRI and AJAMRI charge $200 for papers with "five or more non-Indian authors"; AJFMR charges $80 for one to four authors and only $10 for five or more authors.
The editor-in-chief of all three journals is listed under the same name: Denish Brown. The contact is a WhatsApp number with a US area code. The editorial boards list dozens of members, most from Indian universities.
But this editorial board list also does not hold up to scrutiny. Kanakaraju Manupati, a postdoctoral researcher at Virginia Commonwealth University, is among them. He told THE: "I was unaware of the existence of this journal and have no association with it. I just visited the journal website and was surprised to see my name on the editorial board."
In other words, from authors to editorial board members, the "academic endorsements" of these journals are very likely stolen wholesale. THE contacted the three journals and had received no reply as of publication.
Manning: "This is an even worse scam"
After the report was published, Christopher Manning, who was named, responded on 𝕏. He had previously seen two fake papers "attributed" to him, published in the so-called International Journal of Computer Vision and AI Applications and International Journal of Clinical and Medical Sciences, respectively.
The former's name is obviously riding on the coattails of the top computer vision journal IJCV (International Journal of Computer Vision); the latter is even more absurd — an NLP scholar "published" in a clinical medicine journal. It is worth noting that these two journals are not among the three investigated by THE, which means similar operations may involve far more than one gang.
"Scam journals have always existed; this is a nastier scam," Manning wrote. "Most important is to keep them from being indexed! Please, Google Scholar!"
In the comments, a user complained that Google Scholar kept recommending obviously fake papers to them, with a bunch of big names listed in the author column. "I really can't figure out the motive — why would anyone do this?"
Manning's reply pinpointed the key: THE's report believes that these journals list well-known scholars as authors in order to attract PhD students to submit papers too, and these journals charge fees. He added one last line: "It might actually work!"
The business of impersonation: using big names as billboards
This is exactly what Said worries about most: "If a newly enrolled PhD student sees my name in this journal, he might think submitting here is fine too. I really don't want that to happen."
Traditional predatory journals rely on "fast publication, accept everything submitted" to attract authors desperate to pad their publication counts. But they generally share one weakness: they don't look like legitimate journals. Generative AI happens to fill this gap — in minutes it can mass-"write" papers with standard formatting, complete citations, and an appearance of substance, then attach the names of top scholars, and a predatory journal can "package" a respectable back catalog and a luxurious author lineup within weeks.
This playbook is not new. Diomidis Spinellis, a software engineering professor at the Athens University of Economics and Business, experienced exactly the same thing in December 2023: a paper titled "Global Business Strategies in the Digital Age" was published under his name and affiliation in the Global International Journal of Innovative Research (GIJIR), complete with a valid DOI.
Spinellis did not stop at clarifying. He crawled the entire GIJIR website for analysis and published a case study in Research Integrity and Peer Review in 2025. Of the 53 articles with the fewest citations, at least 48 appeared to be AI-generated, with Turnitin detection giving AI probabilities as high as 100% on multiple papers. His conclusion: these journals use impersonated AI papers to raise their own status on the one hand, and on the other to attract authors who want to pad their publication counts.
According to a JMIR report in June this year, Spinellis classified GIJIR papers into three categories: the majority are "fake-attribution papers" generated by the journal itself using AI and falsely attributed to scholars; next are "purchased-attribution papers" bought by authors with money; genuine research is actually the minority.
Similar incidents clearly became more frequent in 2026. In April this year, the well-known cardiologist Eric Topol discovered that his name had been used by a digital health journal; in May, NBC News reported on a network of fake journals that had published hundreds of AI-generated papers, some of which likewise falsely used the names of professors at top universities.
The real vulnerability is downstream
If these papers merely sat on an obscure website, the harm might be limited. But Manning's appeal points to the real crux of the problem: indexing.
Said discovered the paper precisely because Google Scholar had included it on his homepage and actively pushed it to him. Once indexed by academic search engines, these papers appear on scholars' personal pages, enter search results, and may even be cited by other papers. Spinellis's impersonated paper garnered 3 citations within months of publication.
For the AI industry, there is a more insidious risk: a large amount of model training data comes from crawled academic literature. When AI-generated junk papers are attached to real scholars' names and enter indexes, they may flow back into the training corpora of the next generation of models, forming a pollution loop of "AI-generated, AI-indexed, then learned by AI again."
Technically, this is not an unsolvable problem. The directions Spinellis proposed in his research include strengthening identity verification mechanisms for DOI and ORCID, improving AI content detection, and reforming a research evaluation system oriented around paper counts. The JMIR report also mentioned that before submitting, researchers can use checklists such as Think.Check.Submit to verify a journal's ISSN, DOI, and independent indexing status.
But all these measures cannot get around one reality: the cost of generating a fake paper has approached zero, while the cost of clarifying a fake paper falls entirely on the victimized scholars. After Said complained, the paper switched to a different author and stayed up; Leskovec only learned he had "published" a paper when a reporter emailed him; Manupati had no idea he was listed on the editorial board.
So here is the question: when names like Fei-Fei Li and Manning can be borrowed at will, to what extent can academic authorship itself still serve as a credential of trust? In an era when AI makes "writing papers" cheap, what academia needs to protect may no longer be just the authenticity of content, but also the authenticity of names.
Reference links
This article comes from the WeChat public account "Machine Heart" (ID: almosthuman2014), author: someone concerned about AI abuse, published by 36Kr with authorization.