The Propulsion Papers—Part II: Analysing the anatomy of a Hijacked Journals and Fake AI Citations

To understand how this happened, we need to go beyond Fake AI Citations and analyse the artefact she found. I feel a little conflicted over naming the specific piece, but I chose to use it as a case study. It is not my intention to punch down, and I will address this issue later. The paper in question appears to be a standard academic PDF. It has an ISSN and a has a volume number for the Tuijin Journal it proports to belong to.

Why Propulsion? Are Orwell’s pigs flying now?

But it is the fruit of a Hijacked Journal. This may be a new term for some people, and indeed I only encountered it myself recently in my capacity as one of the editors-in-chief of the Journal for the Psychology of Language Learning (JPLL). In an editorial meeting we were discussing another journal which had been hacked recently, their editors locked out, their ISSNs and DOIs stolen and appropriated by a predatory journal.

The mechanics of this digital identity theft are as mundane as they are devastating. It usually begins with a simple administrative slip-up—a missed renewal email, an expired university credit card, or a momentary lapse in domain registration by a small, underfunded academic society. The very second that URL expires, automated sniper bots operated by international scam networks snap it up. In an instant, the legitimate editors are locked out of their own digital house.

The scammers do not just steal the name; they clone the entire architecture of the website (Abalkina, 2021). They slap the stolen ISSNs, real editorial board names, and DOI prefixes across the homepage to create a perfect, hollow replica. For the real editorial board, reclaiming their hijacked identity becomes a Kafkaesque nightmare. They are suddenly forced to wage a bureaucratic war against anonymous hosting providers, domain registrars, and ICANN to prove they are the rightful owners of their own journal. This process can drag on for months.

While the real editors are drowning in paperwork, the hijackers are open for business. They actively solicit submissions, charge extortionate Article Processing Charges (APCs) of $100 to $500 to desperate researchers, and pump out hundreds of unreviewed, garbage PDFs under the journal’s hard-earned reputation. These are not the obvious spam emails of the past; these are sophisticated phishing operations that wear the stolen skin of legitimate science (Abalkina, 2023). By the time the real editors manage to get the scam site taken down or migrate to a new web address, the damage to their academic legacy is already done, and the scholarly water supply has been thoroughly poisoned. This poisoned well is very much part of the root-cause of Fake AI Citations.

Previously, predatory journals had simply been those which charge extortionate fees and send unsolicited submission requests; most of them end up in the spam box of your email. In fact, you know you’ve finally ‘made it’ as an academic when you stop getting excited and instead you are simply cynical whenever you get a flattering email about you being an ‘expert in your field’.

Luckily for me, my journal has not been hacked and, as far as I know, my words have only been sold to AI when I was in the loop. This was more than just a case of Fake AI Citations. So how had this Hijacked Journal’s content ended up on my marking pile? Before we dissect the text, we must define the vessel.

But how does a brand-new fake journal immediately convince researchers it has been operating for years? It builds a “Zombie Archive”.

As Abalkina (2021) outlines in her research on hijacked journal networks, a newly minted clone website has a problem: an empty archive won’t attract paying victims because it lacks the illusion of an established, continuous publication history. To solve this, the hijackers don’t bother writing original fake papers—that is far too expensive. They also avoid stealing directly from legitimate, high-profile journals, as that invites immediate copyright strikes and easy detection.

Instead, they rely on a massive, incestuous ecosystem of recycled text. Research by Dadkhah et al. (2016) demonstrates that identical texts constantly circulate between various hijacked and predatory journals. Because these bottom-tier predatory papers are practically invisible to the wider scientific community—with Björk et al. (2020) finding that 56% of articles in predatory journals receive zero citations—the hijackers can safely scrape and reuse them without anyone noticing. Again, this is part of the mechanics of why LLMs tend to hallucinate Fake AI Citations.

The scammers operate in organised networks, shuffling a single set of plagiarised texts between dozens of different clone websites like a deck of cards. This perfectly explains why an essay on Animal Farm suddenly appeared in a journal dedicated to rocket propulsion. The scammers did not care about the discipline; they simply needed raw, untraceable text to instantly flesh out the fake archive of their latest hijacked domain.

For a more detailed discussion of this phenomena, I recommend the fantastic work of Anna Abalkina, who has extensively analysed the methods and networks used by these nefarious actors. Also, Medhi Dadkhah has covered this topic quite extensively, building on the foundational whistleblowing of Jeffrey Beall. He is the librarian who originally sounded the alarm on predatory open-access publishing and created the infamous (and now archived) Beall’s List, which is where I went to confirm the status of the Tuijin Jishu (Journal of Propulsion Technology) which my student had cited and which caused me all this extra work!

Most alarmingly, Abalkina (2023) reveals that Scopus—the barometer of academic prestige, aka the annoyingly too-cool-for-school kid who chooses who gets to be legitimised or looked down upon in academic fields—is deeply compromised by a practice known as indexjacking.

Indexjacking is the systematic infiltration of hijacked journals directly into major international indexing databases. Scammers don’t just fool desperate authors; they fool the institutional gatekeepers. They do this by compromising the homepage link in the legitimate journal’s Scopus profile, redirecting all official traffic to the scam site, or by managing to get their unauthorized, unreviewed slop indexed directly into the database alongside legitimate research. Abalkina identified at least 67 hijacked journals that have penetrated Scopus in this way.

This is where the system completely fails us. Scopus operates on a sluggish, purely reactive policy. When they finally notice a journal has been hijacked—often months or years later—they simply delete the compromised profile or unauthorized content quietly, without any transparent public documentation or formal retraction notices. But by the time Scopus sweeps the mess under the rug, the damage is terminal. That unreviewed slop has already been cited by other papers, and the metadata has been automatically exported downstream into other critical, trusted databases, like ORCID profiles. The database itself becomes a laundering machine for fake science, granting the ultimate stamp of legitimacy to algorithmic garbage.

This is why an analysis of Animal Farm appears next to schematics for solid-fuel boosters.

If we strip away the absurdity of the journal title and look at the text itself, the black hole deepens. This document is a perfect case study in the Ouroboros of modern content production—a snake eating its own tail.

1. The Forensic Evidence of “No-Review” This paper proves that no human eyes reviewed it before publication.

Beyond the glaring mechanical errors, the actual prose of the paper is a masterclass in hollow, algorithmic filler. The text relies heavily on the kind of generic “bot-speak” that plagues the modern content-mill ecosystem, leaning on tired, boilerplate phrases like “vividly illustrates” and sweeping generalisations about the “cyclical nature of political revolutions” without ever engaging in rigorous, close textual analysis. There is no genuine literary critique happening here; it reads entirely like an empty word-count exercise, churned out solely to hold up a title and secure a DOI in a broken “publish or perish” economy.

There is a literary perfection to this failure. The themes of Animal Farm map disturbingly well onto the hijacked journal system.

The ultimate casualty in this Ouroboros of slop is authenticity. When the boundary between a rigorous, peer-reviewed human insight and an algorithmic hallucination is erased by the very institutional databases designed to protect it, authenticity ceases to be a valued metric. It becomes an active liability in a system engineered entirely for volume and velocity.

I am looking at this offending paper, the titular Propulsion Paper; Orwell’s rocket powered flying pigs. While it is easy to laugh at the Rocket-Pig incongruity, there is a human tragedy here. We shouldn’t be punching down at the author of the Propulsion Paper. If this author is indeed a real person—and not just another hallucination in the machine—she represents one of two bleak realities in modern academia.

The first reality is that she is a desperate researcher trapped in a neo-colonial “publish or perish” grinder. According to Abalkina’s demographic analysis, the vast majority of authors publishing in these hijacked vessels come from lower-middle-income countries. Scholars in regions like India, Iraq, and Uzbekistan are often subjected to brutal institutional mandates requiring Scopus-indexed publications for graduation or tenure. This creates a desperate, captive market that scammers specifically target and exploit for exorbitant Article Processing Charges.

But given the old volume number attached to this specific paper, there is a second, even darker possibility: The author might have no idea this PDF even exists.

As we saw with the Zombie Archive mechanics, scammers actively scrape unindexed university repositories, local non-English journals, and undergraduate theses to instantly populate their fake websites with raw text. The author might simply be an undergraduate student who uploaded a perfectly innocent Animal Farm essay to her local university’s intranet five years ago. Now, a cybercrime syndicate has stolen her intellectual property, slapped a fake DOI on it, and used it as structural filler to prop up a hijacked rocket science journal. Her crime of having a work listed in ALL CAPS in her citations pales to nothing when we look at the crime that found the paper propelling into an indexed academic journal.

Whether she is an exploited academic who paid a predatory fee she likely couldn’t afford, or an unwitting student involuntarily drafted into a cybercrime syndicate’s façade, she is a victim of the exact same systemic pollution that is confusing my own students. The villain here is not the student, nor the writer; it is the broken metrics system that created the demand, and the hijacked journals that supply the slop.

The author claims Orwell illustrates the “cyclical nature of political revolutions”. How fitting that this paper ended up in a journal about Propulsion. It proves that the only thing we have truly perfected is the efficient, high-velocity delivery of bullshit.

Before we wrap up this autopsy, we need to address the algorithmic elephant in the room. It is incredibly tempting to blame ChatGPT (or any other LLM) for this entire mess. To blame this on Fake AI Citations. But let’s be perfectly clear: this is not an AI problem. AI is merely the accelerant; it is speeding up the scraping and the automation, but the foundation of this scam was laid long before Generative AI became a household name.

As Jalalian and Dadkhah (2015) documented in their history of the phenomenon, cybercriminals registered the first domains for hijacked journals back in August 2011. By 2014, these bad actors had escalated to “mass journal hijacking” operations, manually creating fake websites for dozens of publications, going so far as to target elite publications like the Journal of the American Medical Association. They built this fraudulent infrastructure by hand.

So, who is the real enemy here? It isn’t the writer, and it isn’t the robot. It is the system that created the hijackers.

The academic publishing industry—basically all of science—was already corrupt long before the scammers arrived. The modern rot traces its roots back to publishers like Robert Maxwell, who realised he could get fabulously rich by monetising scientific legitimacy, gating publicly funded research behind extortionate paywalls, and inventing a manufactured scarcity of prestige based on metrics. It has long been an open secret, extensively discussed by scholars like Ken Hyland, that the industry is broken.

The Fake AI Citations, scammers and hijacked journals, predatory publishers and all these other demons look like the obvious bad guys to this story. But it is not them that will break our current models of academia; they simply realised they could sell counterfeit tickets to an already rigged casino. In a way, these hijacked journals are merely the early flies bursting from an already maggot-infested corpse. The system’s obsession with impact factors and the inherent snobbery of scientific knowledge built the exact conditions for this cybercrime to thrive. The ivory tower is blackened with soot and besieged by zombies, and the publishing conglomerates are the ones who left the gates wide open. I will be delving into this issue in later articles.

We have diagnosed the disease (Part I) and identified the pathogen (Part II). But I’m not just here to complain. I still have to teach class on Monday.

So, how do we fix this? If this is a problem that goes even deeper than the massive sinkhole of Fake AI Citations, How do we teach students to survive in a library that is currently on fire?

In Part III, I’m sharing the “Forensic Flip”—how I am turning my classroom into a live citation audit workshop to teach students how to distinguish between fake AI citations and hijacked journals. Sign up below so you don’t miss the solution.

Abalkina, A. (2021). Detecting a network of hijacked journals by its archive. Scientometrics, 126(8), 7123-7148. doi:10.1007/s11192-021-04056-0

Abalkina, A. (2024). Challenges posed by hijacked journals in Scopus. Journal of the Association for Information Science and Technology, 75(4), 395-422. doi:https://doi.org/10.1002/asi.24855

Björk, B.-C., Kanto-Karvonen, S., & Harviainen, J. T. (2020). How frequently are articles in predatory open access journals cited. Publications, 8(2), 17.

Dadkhah, M., Maliszewski, T., & Lyashenko, V. V. (2016). An approach for preventing the indexing of hijacked journal articles in scientific databases. Behaviour & Information Technology, 35(4), 298-303. doi:10.1080/0144929X.2015.1128975

Jalalian, M., & Dadkhah, M. (2015). The full story of 90 hijacked journals from August 2011 to June 2015. Geographica Pannonica, 19(2), 73-87.