Go back

Rise of AI stokes fears research misconduct could accelerate

RPN Live: Call for more action from publishers, funders and institutes to bolster research integrity

Research-integrity experts have called for more action to be taken on research misconduct, amid concerns that artificial intelligence could accelerate a worsening situation.

At a Research Professional News Live webinar held on 18 July, speakers warned of research integrity being undermined by a rise in organised attempts to manipulate the scholarly publishing system.

Elisabeth Bik, a science-integrity consultant known for her extensive work in identifying image manipulation in publications, said problems with research misconduct are “getting worse”.

“What we’ve seen in the last couple of years is organised networks really trying to game the metrics, trying to produce fake papers,” Bik said. She added that “this is a pretty recent and increasingly growing problem”.

Bik said such networks, often referred to as paper mills, are “probably already using artificial intelligence to create fake papers and it’s going to be increasingly hard to detect those”. She suggested that the use of AI would cause research misconduct to increase in scale.

Need for speed

In 2014-15, Bik examined more than 20,000 papers for image manipulation, flagging 782 of them to journals as problematic. But Bik said journals “seem to be very slow to respond to these potential problems”.

As of June 2023, 65 per cent of the papers she flagged had not been corrected or retracted. “I feel, as an editor or journal publisher, you have the obligation to alert the reader that there’s a big potential problem with a paper,” she added.

Sabina Alam, director of publishing and integrity at academic publisher Taylor & Francis, agreed that problems with research misconduct were getting worse.

“I’ve been doing this since 2008 and there’s been a marked difference, I would say, from 2017 onwards,” Alam said. She added that investment in technology was needed to detect fake papers generated by AI.

As well as paper mills, Alam said “cartels” of people were “working with each other to boost their publication numbers and h-indexes”—a measure of the significance of an individual researcher’s published research.

Alam said that the investigation of reports of misconduct and subsequent remedial action needed to be “a lot faster”, but that “it can’t be publishers who resolve this on their own”.

Institutes and funders

Nandita Quaderi, editor-in-chief of Web of Science*, said research institutes, funders and publishers all have responsibility for ensuring research integrity.

“In terms of misconduct, it does have to primarily sit with the institutes, but I think funders also need to play a role,” Quanderi said. She called for more pushback from funders when misconduct is identified.

Marcus Munafò, chair of the UK Reproducibility Network and associate pro-vice chancellor for research culture at the University of Bristol, said institutions “need to own this”, but that problems could arise when they “mark their own homework”.

He called for more transparency regarding research misconduct, saying that once it is established that misconduct has occurred, “that needs to be made clear, and often it isn’t”.

“Often these people who engage in misconduct quietly leave by the back door, maybe join another institution—it’s very opaque,” Munafò explained.

He said that the move toward open research and greater transparency “allows for greater scrutiny”, which is beneficial for research integrity more broadly, since transparency “then creates an incentive for researchers themselves to engage in quality-control processes”.

‘Becoming more visible’

Not all of the speakers were convinced that problems with research integrity were worsening.

Miles Padgett, a member of the UK Committee on Research Integrity and interim executive chair of the Engineering and Physical Sciences Research Council, said that while it was unclear “whether it’s getting worse or better, it’s certainly becoming more visible, and essentially I think that visibility is a good thing”. He added that the move toward open data would help ensure the integrity of the research record.

Padgett also said that it was not enough to agree that there were problems with research integrity, but that “what the agenda needs is understanding what we are actually going to do”.

Debra Schaller-Demers, senior director for research integrity and compliance at New York University, also said that there were not “necessarily…more incidents of research misconduct or breaches of integrity, but I do think we have the tools to spot them more easily”.

She said problems often occurred due to the pressures placed on younger researchers.

“They’re [juggling] a lot of balls in the air; I don’t think that it’s necessarily a deliberate attempt to circumvent rules,” Schaller-Demers said.

Listen to the webinar recording here.

*Web of Science is a Clarivate product. Research Professional News is an editorially independent part of Clarivate