The Rise of ‘Ghost Journals’: How AI-Generated Papers are Hijacking Academic Identities
Table of Contents
A New Frontier in Academic Fraud
The scientific community is grappling with a sophisticated evolution of the ‘predatory publishing’ model. While fraudulent journals have existed for years—charging authors fees to publish papers without legitimate peer review—a new, more insidious trend has emerged: journals that create the content themselves using generative AI and attribute it to established scholars without their knowledge.
These ‘ghost journals’ are not merely skipping the peer-review process; they are automating the entire pipeline of academic production. By leveraging Large Language Models (LLMs), these entities generate plausible-sounding research papers and then scrape the names, affiliations, and expertise of real professors from university directories and existing publications to serve as the authors.
The Mechanics of AI Impersonation
Unlike traditional identity theft, which targets financial data, this operation targets intellectual authority. The process typically begins with the AI generating a paper on a topic that aligns with a specific professor’s known field of study. Because the AI can mimic the technical jargon and structural conventions of a scientific paper, the resulting documents often pass a superficial glance, appearing legitimate to those unfamiliar with the author’s specific voice or current research trajectory.
This creates a dangerous feedback loop. When these papers are indexed in automated databases, they may appear in search results, potentially misleading other researchers or policy-makers. For the professors involved, the discovery is often jarring—finding a fully formatted research paper attributed to them in a journal they have never heard of, containing data they never collected and conclusions they never drew.
Beyond Simple Hallucinations
The core of the problem lies in the bridge between AI’s ability to synthesize information and the inherent trust placed in academic credentials. When a human author uses AI to ‘polish’ a paper, the responsibility for the data remains with the human. In these fraudulent journals, there is no human author; the AI is both the writer and the ghostwriter, with the real professor serving as a convenient, high-authority mask.
This goes beyond the well-documented issue of AI ‘hallucinations’—where a chatbot invents a fake citation. In this scenario, the AI is creating an entire fake persona for the paper, backed by a real-world identity. It weaponizes the perceived prestige of tenure-track positions to give artificial credibility to nonsense or fabricated data.
The Structural Vulnerability of Peer Review
The crisis highlights a systemic vulnerability in how academic prestige is tracked. Most digital repositories and indexing services rely on name and institutional affiliation, which are public record. There is currently no universal, cryptographically secure method to verify that the person whose name is on a PDF is actually the person who submitted it.
As these AI-generated papers flood the ecosystem, the signal-to-noise ratio in scientific literature is deteriorating. The risk is not just an embarrassed professor, but the potential for fabricated medical or technical data to be cited in genuine meta-analyses, leading to a ‘pollution’ of the global knowledge base.
Academic institutions are now urging faculty to set up Google Scholar alerts for their own names and to report fraudulent journals to registries like the Directory of Open Access Journals (DOAJ). However, as AI tools become more adept at mimicking specific academic styles, the line between a legitimate collaboration and a synthetic impersonation is becoming dangerously thin.
