To use this site, please disable the ad blocking feature and reload the page.
This website uses cookies to collect information about your visit for purposes such as showing you personalized ads and content, and analyzing our website traffic. By clicking “Accept all,” you will allow the use of these cookies.
Users accessing this site from EEA countries and UK are unable to view this site without your consent. We apologize for any inconvenience caused.
Reuters file photo The abbreviation for artificial intelligence is seen in this photo.
By Yuki Nagasawa and Takushi Kuwahara / Yomiuri Shimbun Staff Writers
17:47 JST, November 21, 2024
Fake academic papers purported to have been written by Japanese researchers were published on an overseas academic journal website, The Yomiuri Shimbun has learned. The papers may have been written using generative AI, according to an expert.
The website is believed to be a so-called predatory journal that charges authors a fee for publication and usually carries papers without checking them for quality or legitimacy. The expert consulted by the Yomiuri suspects that the website’s operators created the fake papers using generative AI, to make the site look like it represents a quality academic journal.
Kazumichi Fujii, 43, is a senior researcher at the Forestry and Forest Products Research Institute in Tsukuba, Ibaraki Prefecture. Earlier this month, Fujii found two academic theses on the overseas academic journal website that named him as the author. However, he had nothing to do with the papers.
Fujii specializes in soil research and won the 2019 Kawai Hayao Prize for social science and humanities for a book he wrote. He said the fake papers were written in English on a theme that he had studied in the past, but that his alleged affiliations mentioned in the papers were nonexistent organizations at the University of Tokyo and Nagoya University.
Fujii e-mailed the operator of the website and asked for the papers to be deleted, but received no response. On Friday, however, the name of the author on one of the papers was changed to a different Japanese name, although the claimed affiliation remained the same.
The newly added Japanese name was not on the list of faculty members at the university mentioned on the website.
Fujii said he has never been contacted by the website’s operator.
This website is one of about 1,300 academic journal sites that have been cited by overseas researchers as possible predatory journals. It carries a large number of papers and offers to publish theses for a fee.
The alleged authors named on the website include one believed to be a Japanese person other than Fujii. Other papers said to have been written by people at research institutions in Japan and overseas were also found.
It is not clear why fake papers were carried on the website under Fujii’s name. But the website may have been trying to pass itself as a high-quality website by publishing the papers said to be written by established researchers, in order to solicit fees in exchange for publishing research papers.
“There are no errors in the factual content of the thesis being carried on the website under my name. But it does not offer any new findings, and it annoys me when people think I wrote it,” Fujii said. “This is an unforgivable act that damages the credibility of scientific papers.”
Isao Echizen, a professor at the National Institute of Informatics and an expert in information security, said both of the two papers “are likely to have been created using generative AI.” Echizen analyzed the papers at the request of The Yomiuri Shimbun.
A researchers group led by Echizen developed a system to determine whether English-written papers were created using generative AI. The system learned from a large amount of papers written by researchers and those made with AI, and can detect differences in the writing that humans cannot pick up on, he said.
“The papers in question have strange structures as theses. They lack citations, for instance,” Echizen said.
“Examining each sentence, however, they don’t look that unnatural. There are fears that students and other people who are not familiar with these matters may mistake these papers for real ones,” he said.
“The harm caused by fake papers bearing the name of existing researchers as authors has been reported overseas, and the damage may have been spreading further, unbeknownst to Japanese researchers,” said Eisuke Enoki, who heads the Kaseiken study group on science, policy and society.
“With the advent of generative AI, it’s become easier to fabricate papers. Fears are growing over the contamination of the academic world,” said Enoki, an expert on research ethics.
