Journal vanishes after listing Eric Topol and others as authors of a fabricated AI paper

Researchers at several institutions learned in April that their names appeared on a paper they had not written, published in a journal they had never heard of. Within a day, the journal and its publisher had disappeared from the web.

Cardiologist Eric Topol, executive vice president of Scripps Research and founder of the Scripps Research Translational Institute, posted on X on 22 April that his name was on a "fraudulent" paper in the Journal of Digital Health Implementation, according to Retraction Watch. He told the site that the article, dated 29 March and titled "Implementation Science for AI Integration in Digital Health Systems," was probably AI-generated. "If there ever was an AI-generated paper, this one would qualify as a high probability of being so," he said.

Michael Matheny, a professor at Vanderbilt University Medical Center, found out when a colleague asked him about the article. He told Retraction Watch he alerted the other listed authors after spotting "small errors in the naming and affiliation," including a wrong middle initial for himself and wrong first initials for several others.

Five of six listed authors say they never saw the paper

Of the six names on the paper, five told Retraction Watch they had never seen it before. The sixth, listed as A. B. Martin with the Imperial College of London, could not be matched to any researcher. One of the real names belonged to Effy Vayena of the Swiss Federal Institute of Technology in Zurich, who said the paper "clearly used names of well-known academics which increases the likelihood that the wider community might want to read it or potentially cite it." Topol is among the ten most cited researchers in medicine.

Topol said adding the authors' names was "grossly illegal." Vayena framed the problem more broadly. "What scares me is that we've reached a point where it's difficult to protect our academic work and identity," she told Retraction Watch. "This is incredibly dangerous not only for individual reputations and trust in academic work but also for the actual scientific knowledge."

The infrastructure behind the journal was thin. According to Retraction Watch, the publisher, Ellinger Publishing Media, claimed to be a UK company but does not appear in the UK business registry, and its listed address is a coworking space in London. Before removal, its homepage resembled Springer Nature Link, the portal for that publisher's titles. Ellinger listed four journals in total, and the issue carrying the fake article was the first volume for both the journal and the publisher.

"This is incredibly dangerous not only for individual reputations and trust in academic work but also for the actual scientific knowledge." Effy Vayena, Swiss Federal Institute of Technology in Zurich

The paper's DOI pointed to Zenodo, the open repository, and that page was taken down for copyright infringement. Archived versions of the top-level domain, e-pubmed.co.uk, show it was previously tied to a Russian-language publication.

Identity theft in academic publishing is not new. Researchers get impersonated as reviewers or former colleagues, and hijacked journals borrow the credibility of established titles. Retraction Watch notes that some journals and organisations have floated identity checks as a fix. A commenter on the Retraction Watch piece pointed to opensci.id, which verifies ORCID identifiers against government IDs, and said the digital health publisher JMIR Publications is about to pilot the system for reviewers and selected guest editors and authors.

For digital health research, the timing is awkward. The field is publishing fast on AI in clinical settings, and a fabricated paper with famous names attached could have been picked up and cited before anyone checked. Topol's post went up in the early afternoon. By the evening, the site no longer loaded.