Several Chinese universities have launched retrospective reviews of previously published academic papers to strengthen research integrity across the academic system and tackle longstanding problems such as data fabrication, image duplication, improper authorship and industry chains that produce and sell academic manuscripts — known as paper mills.
Guangdong Polytechnic requires researchers to review papers formally published since Jan 1, 2017, and cross-check databases including the China National Knowledge Infrastructure and Web of Science for retractions, corrections, data fabrication, image duplication and paper trading, according to a notice recently issued by the university.
Researchers must also verify author order and institutional affiliations, sign integrity commitment letters and report problems truthfully, the notice said.
Anhui Jianzhu University and Anhui College of Traditional Chinese Medicine issued similar notices covering papers published since Jan 1, 2020.
Anhui Jianzhu University focused on image duplication, improper authorship and paper trading, and required researchers to check whether data was accurately recorded and experimental processes were traceable. It also issued rules on presubmission review and data retention, requiring ethical review for studies involving human or animal subjects.
The notice issued by the Anhui College of Traditional Chinese Medicine lists eight categories of priority problems, including plagiarism, duplicate publication, and falsification or repeated use of image data.
In recent years, China has tightened requirements for research integrity, requiring universities not only to investigate and handle academic misconduct but also to establish routine supervision, process management and risk prevention mechanisms.
In 2024, the Ministry of Education issued a guideline requiring universities to take primary responsibility for research integrity, establish specialized bodies, coordinate across departments, and incorporate integrity checks into teacher recruitment, promotion and assessment.
Universities have also continued to publicly disclose confirmed cases. Sun Yat-sen University and Nankai University, for example, have published notices involving image and data problems and authorship issues. Those responsible received penalties including removal from office, demotion, disqualification from promotion and project applications, and termination of employment.
Tang Daisheng, a professor at the School of Economics and Management at Beijing Jiaotong University, said the reviews aim to verify the authenticity of existing research and authors' contributions, purify the academic ecosystem and support a shift away from overemphasizing papers in academic evaluation.
Research integrity is a bottom line that should not expire, Tang said, adding that retrospective checks must be combined with presubmission review and data retention to form a closed loop of governance.
The deeper gap being addressed, Tang said, is the longstanding neglect of postpublication checks, as published work is often mistakenly treated as exempt from inspection.
He cited the Law on Academic Degrees, which allows degrees to be withheld or revoked in cases involving ghostwriting, plagiarism or fabrication, as evidence that accountability for misconduct has no time limit.
Tang called for a long-term mechanism centered on four priorities: consolidating institutional responsibilities across departments; embedding research integrity into teacher management through integrity files and commitment letters; strengthening presubmission review and data retention; and reforming academic evaluation to break the "papers only" approach.
"Only in this way can papers return to their original purpose of seeking truth," he said.
A second-year journalism graduate student at Sun Yat-sen University surnamed Zhang said she feels pressure from the reviews and is particularly concerned about self-reviews of degree theses, noting that she has seen cases in which master's degrees were revoked.
She said she plans to follow the university's requirements on controlling AI-generated content and retaining original materials, although she finds keeping them after graduation cumbersome.
The reviews can serve as a deterrent, Zhang said, especially as AI tools are increasingly capable of leaving identifiable traces, making it harder to pass off AI-generated papers as original work on important occasions.
Yu Xinyan contributed to this story.
