
What happened
Ayoolewa Aregbesola and Sola Emmanuel Owolabi published a Z-Library paper on October 6, 2026; TorrentFreak found at least 10 of its 44 references inaccurate, including a TorrentFreak article it never published.
Why it matters
The paper's apparent sourcing is unreliable on its face, which may weaken its findings on shadow libraries and access in resource-limited regions.
WHO IT HITSAcademic publishers and journal editors who screen citation integrity, and researchers who rely on Z-Library access studies, may need to reassess this paper.
Summaries like this, every morning.
The paper by Ayoolewa Aregbesola and Sola Emmanuel Owolabi set out to examine shadow libraries like Z-Library, noting that they improve access for researchers in resource-limited regions while raising ethical and legal concerns. In a chapter on legal action against Z-Library, the paper describes major academic publishers Elsevier and Springer Nature as taking similar legal steps, and it supports that with a 2021 TorrentFreak article. TorrentFreak says no such article was ever published.
Beyond that missing citation, TorrentFreak's review of the full reference list found many entries that look plausible because they cite real publications or organizations in context, yet are not accurate. It points to a 2019 College & Research Libraries item credited to Carol Tenopir: she has published on similar themes, but the cited issue does not contain an article by her. Of the 44 references, at least 10 were inaccurate in this way.
The paper's acknowledgments state that the literature search AI tool Consensus AI was used to review a large volume of literature and extract relevant information, and that ChatGPT was used to improve wording in parts of the paper. TorrentFreak notes that if AI-found papers had been scrutinized, reference errors would not occur, suggesting AI may have been used for the whole writing process rather than as an aid. It also reports that its analysis of the paper's abstract and conclusion with the AI detection tool Pangram found both sections were likely 100% AI-generated. Taylor & Francis has said it will investigate.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
At an IMF meeting in Bangkok, cofounder Jack Clark said this year brought convincing evidence the singularity…

Anthropic filed an S-1 for a Nasdaq listing under the ticker ANT, disclosing revenue of more than $16 billion…

Overwatch, co-founded by Arjun Bisen and Zara Perumal, raised a $15 million Series A led by Mouro Capital, wit…

The operator divided Claude Code permissions into three tiers in CLAUDE.md: act without asking (code edits, co…

Resolution used an orchestrator agent to spin up sub-agents that wrote Lean proofs and did pen-and-paper math…

The biggest problem with AI wipeout warnings is that they rarely explain how a rogue model would actually pull…
