Expert perspective: Wu Guangheng Looks at the Future of Research Integrity

In late February we gave a presentation on the Future of Research Integrity at the Researcher to Reader Conference in London. In preparation we sought opinions and insights from leading research integrity sleuths, many of whom were generous with their time and expertise. Among these was Wu Guangheng, Ph.D, founder president of the 5GH Foundation, a non-profit organisation dedicated to promoting responsible science and technology research. The foundation has conducted numerous investigations into suspected academic misconduct. 

This is an edited version of insights Dr. Wu shared with Signals, shared with the permission of Wu and the 5GH Foundation.

1. Should individuals who identify and report problematic research receive more recognition or rewards? If so, what form could that take?

Academic sleuths who identify and report problematic research practices are not only vital contributors to the scientific community, but to society as a whole, and they deserve recognition for their contributions. However, academia heavily relies on publication and citation metrics to evaluate researchers, especially those at early career stages. Young researchers are not incentivised to devote time to identifying problematic research practices or pursuing careers as academic sleuths.

It is time to start talking about a metric to evaluate the contributions academic sleuths make in identifying problematic research practices, which could be used in considering promotions and funding. While any metric has its limitations, an open discussion on this matter could bring significant benefits to our scientific community.

2. Looking ahead to the next three years, what threats to research integrity worry you most?

The development of artificial intelligence (AI) may make fabricated content, such as spectra and data, more difficult to detect. Dr. Mu Yang with Columbia University has identified tens, if not hundreds, of hand-drawn and/or photoshopped spectra since the summer of 2024 (example here). Her work, and that of other sleuths, will become much more challenging if authors use more sophisticated technologies to fabricate spectra. The 5GH Team is working with our partners to examine methods for generating false spectra, in the hope of identifying key features that can be used to detect them. 

Other forms of research misconduct are also likely to evolve with the advent of new

technologies. For example:

• Plagiarism is no longer limited to copy-pasting and tortured phrases are no longer necessary. Sentences, paragraphs, and even entire articles can be rephrased and reorganized using generative AI tools like ChatGPT and DeepSeek. 
• False questionnaires or statistical tables no longer need to be manually filled out. Data with specific characteristics can be automatically generated.

Under these circumstances, the strategies we previously used to detect false content may no longer be effective. It remains unclear whether journals can prevent such content from being published and whether we can detect it.

3. What steps do you think researchers, institutions, and publishers need to take to improve research integrity?

Academic sleuths receive little to no support from their institutions, which are also generally extremely slow to respond to reports from sleuths, if they take any action at all. We have numerous examples of universities in China that have failed to respond to reports from sleuths after one to two years. So, in addition to taking reports from academic sleuths seriously, institutions should form their own integrity teams and provide team members with the necessary resources, including training programs, technical support, and financial funding. 

Publishers and funders could support more transparency in academic publishing. Some examples where this can make a difference include:

Since 2024, the 5GH Team and our allies have identified over 1,000 cases of “editor-author conflict of interest”, in which journal editorial board members or academic editors handle their own articles or those from frequent coauthors. These conflicts were primarily discovered because editor information was openly disclosed. Dr. M. A. Oviedo-Garcia with Universidad de Sevilla has uncovered large numbers of problematic review practices from the open peer review reports.

Sometimes peer reviewers ask or even force the authors to cite their own articles. Since peer review reports are usually unpublished, academic sleuths have to perform intensive citation analysis to identify potentially problematic reviewers. A startling example of this involved a group of Chinese authors who were forced to cite 13 articles from two reviewers, stating that “as strongly requested by the reviewers, here we cite some references [[35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47]] although they are completely irrelevant to the present work” . Since the publisher did not disclose details of the reviewers concerned in the retraction notice, their names remain unknown to the public, and the question of any editor involvement not addressed. Mandatory guidelines about the publication transparency are urgently needed.

Besides the institutions and publishers, other parties, especially legislative and judicial authorities, should also act to strengthen the research integrity. Few countries have legislation regarding academic misconduct. Prohibitions on selling authorship remain a contentious issue, because laws uphold the right of individuals to transfer their intellectual property. As a result, authorship-for-sale advertisements are widespread on Chinese social media sites like RedNote. Service providers such as RedNote and Facebook show no willingness to remove these advertisements, as they are not illegal under current laws.

4. How do you think research integrity is likely to evolve in the next few years?

Academic misconduct is often seen as the dark counterpart to science, fueled by advanced technologies. We cannot envision a bright future without immediate and coordinated action among the key stakeholders.

Signals is used by sleuths, including Dr Wu, for both research and for sharing and amplifying their work, which are represented as Expert Contributions.