For academics, an allegation of research misconduct can lead to a lengthy and difficult process. The rapid development and growing use of AI are also raising new questions about research integrity, with recent cases in Sweden highlighting some of the challenges involved. Against this backdrop, we examined how universities across the Nordic countries are responding, and what support is available to researchers facing allegations of research misconduct.
Cases involving AI-related errors are becoming more common, according to Jonas Jeppson, Executive Director of the Swedish National Board for Assessment of Research Misconduct, as reported by Dagens Nyheter.
Researchers increasingly use AI for reference management and language editing, but failing to verify its output can lead to errors. Jeppson cited a recent case at Linnaeus University in which four fabricated references were included in a research funding application. Because the researcher had an opportunity to verify them, the Board found that the researcher had acted with “gross negligence”.
Another case discussed in the article, reported by the Swedish media outlet, differed in both nature and outcome. It involved Virginia Dignum, Professor of Responsible Artificial Intelligence at Umeå University. Dignum faced an anonymous complaint alleging research misconduct in connection with her recently published book, The AI Paradox. The allegation rested on two claims: that AI-detection tools had flagged parts of Dignum’s text as AI-generated and that six references were fabricated. The Board later dismissed the complaint.
For Dignum, however, the process itself was isolating, and, in her view, raised questions about how institutions support staff facing such allegations:
– The accusation itself did not shake me; I knew the facts. The process did. Being formally accused of misconduct and then hearing nothing from your own institution is isolating in a way I had not expected, and it takes up mental space that should go towards work. What was missing was humanity: contact, warmth, someone asking how you are. These cannot be added to a checklist, and they are increasingly disregarded in the growing over-checklist-based management of universities.
– I coped by doing what I do professionally: setting out the facts in writing, publicly, in my own words, and leaning on colleagues who know my work.
– That is a privilege of seniority, and I am aware of that.
Dignum also argues that AI-detection tools should not be treated as evidence:
– AI detectors are not evidence. When an anonymous complaint based on their output is enough to start a formal process, the system becomes a tool for harassment, and early-career researchers are the most exposed.
When Dignum posted on LinkedIn about her experience, fellow researchers were eager to join in the conversation, with over a thousand reactions and 173 comments as of September 22.
The response to Dignum’s account on LinkedIn reflected some of the issues she raised. Researchers shared messages of support, questioned the reliability of AI-detection tools – one noted that a paper published in 1999 had been flagged as AI-generated – and criticised what they saw as a lack of institutional support for scholars facing such allegations. Several also pointed to the vulnerability of researchers at earlier stages of their careers, highlighting growing unease about accusations of AI use and the potential consequences they may face.
Uncertainty about Rules and Available Support
Concerns about research misconduct are not confined to social media feeds. They were also raised recently at a Nordicom workshop on academic publishing for early-career researchers, as Kristin Clay, Manuscript Editor at Nordicom, described:
– During my seminar on finding instructions for authors and preparing a manuscript for submission, the topic of AI dominated the conversation among the PhD students taking part. There is clearly some trepidation and uncertainty about when and how to use AI responsibly in the whole process. Specific concern was raised about what they should do if they are accused of using AI inappropriately, with one attendant even saying “it’s a bad time to be a doctoral student right now”.
A recent report by the Swedish Union of Civil Servants and the National Union of Students found that 45% of doctoral researchers are uncertain about permitted AI use, while more than one third say this uncertainty causes work-related stress.
A recent report by the Swedish Union of Civil Servants and the National Union of Students found that 45% of doctoral researchers are uncertain about permitted AI use, while more than one third say this uncertainty causes work-related stress.
To assess whether these concerns are shared within the broader Nordic research community, we contacted media and communication scholars at different career stages across the region. We asked about their concerns regarding allegations of inappropriate AI use, whether they know where to turn for psychological, communications, or media support, and whether they believe their institutions provide clear guidance on how to respond to such allegations. Those who responded were given the option to remain anonymous.
Several respondents weren’t worried about being accused, though they acknowledged the rising importance of the issue. For instance, a Danish researcher wrote:
– I am not so worried about being accused of (mis)using AI tools as I don’t use these tools at all in my work. However, there has recently been a case in Copenhagen about an employee having misused AI where it led to critical errors – so now it might have become top of mind for many.
Concerning institutional support, the same researcher pointed out the University of Copenhagen offers general psychological support and has a committee that reviews general question of scientific (mis)conduct, concluding that “they are probably the primary authority on this topic locally”.
– At the moment, I don’t know exactly who in my organisation I could turn to for support in that exact type of situation. […] I am sure that if these allegations become common place, KU would make further guidelines about this. But it is probably still too new a thing, they stated.
A young Norwegian scholar was also not particularly concerned about being accused of research misconduct, but remained cautious about the use of AI, not least due to unclear institutional guidelines and support:
– I personally use AI very little. I still think it is unclear what the guidelines actually are regarding what is and is not permitted in different contexts, and I want to stay well clear of any potential grey areas. I don’t feel that this is something we are informed about, and it seems like it is something you would have to figure out for yourself if you ended up in such a situation.
The scholar’s comments reflect broader uncertainty among researchers at the University of Bergen about the rules governing AI use and where to turn for support when questions of research integrity arise.
A Swedish professor added nuance to the concerns of allegations:
– I am not concerned about [legitimate allegations] of using AI tools or of engaging in research misconduct. I have, however, thought about the fact that my research area in recent years may provoke certain groups in society, and that they might try to target me – for example, by reporting me for “research misconduct”.
– If I were accused of using AI improperly in my work, I would have absolutely no idea how my department would support me or where I should turn for help, the professor continued. My university does not provide any clear information on this and has not published any guidance for researchers. In fact, I have no idea at all how my institution handles issues of research misconduct or what would happen if someone reported me for suspected research misconduct, regardless of whether AI was involved.
– If I were accused of using AI improperly in my work, I would have absolutely no idea how my department would support me or where I should turn for help.
In Oslo, a professor pointed to the Research Ombud at the University of Oslo as a possible source of support:
– I’m not particularly concerned about this matter. To be transparent, I often include statements explaining how AI has been used in my academic work. The Research Ombud at the University of Oslo would be a natural point of contact, but I think I would first involve the department leadership.
What Support You Can Count On
We also asked the universities about procedures for handling allegations of research misconduct and what kind of support an accused researcher can expect, the latter being also sent to academic trade unions.
Those who responded generally referred researchers to the established bodies responsible for examining suspected deviations from good research practice at their institutions, as well as to relevant research integrity policies and documents.
National guidelines were also referred to, for example, the recently published Artificial Intelligence in Research: Research Integrity and Ethical Principles by the Finnish National Board on Research Integrity.
The Swedish Research Council has Guidelines for the use of AI tools, but only for writing and assessing research funding applications. The Council, along with The Danish Agency for Higher Education and Science, points to the European Commission’s recommendations in the Living guidelines on the responsible use of generative AI in research. While Norway and Iceland several initiatives fostering responsible AI use in general, none were found that explicitly target research.
As for expected support, the majority of our contacts pointed to the faculty and immediate managers as the main contact points. According to Pia Fredberg Nielsen from the University of Copenhagen, “individual researchers are advised to approach their immediate manager, a named person, or the universities’ practice committees when disputes arise”.
Similarly, Jan Lif, secretary of the Council for Examining Deviations from Good Research Practice at the Vice Chancellor’s Office at the University of Gothenburg, pointed out:
– It is the responsibility of the immediate manager (normally the head of department) to promote a good working environment, and in the event that an employee is accused, the management shall take the measures deemed appropriate during the investigation.
– It is the responsibility of the immediate manager (normally the head of department) to promote a good working environment, and in the event that an employee is accused, the management shall take the measures deemed appropriate during the investigation.
Concerning psychological and communication support, the University of Copenhagen, as noted by a respondent above, provides general mental health services, including psychological and psychiatric support, and states that “the Communications Office at UCPH is available to assist researchers who need help managing media-related cases”.
In Finland, Aura Kivilaakso, Head of Development at the University of Helsinki, told us by email:
– The University aims to ensure that researchers have access to appropriate support throughout the process. Researchers may seek guidance from supervisors, HR services, occupational health services and, where relevant, legal advisers. In addition, the University has a network of Research Integrity Advisers. These advisers provide confidential, low-threshold guidance and discussion support on research integrity matters. [… and] remain neutral. […] Where a case attracts public attention, communication support may also be considered on a case-by-case basis.
The University of Stockholm replied:
– There is no formal procedure or structure for this [psychological help and communication support], but in principle such support is available. The challenge is to channel it to those in need in a way that is neither too intrusive nor too passive. Formally, the faculty is responsible for protection and support […]. In practice, researchers involved in cases where allegations are investigated (as reported or reporting parties) have been offered, for example, support by a person proposed by the faculty, psychological support/counselling through occupational health services, media management support from the communications office, and security-related support from the security unit.
Swedish trade union Saco-s reported:
– When a member is accused of research misconduct, the union’s role is to support and advise the member, safeguard their interests, and contribute to ensuring that the matter is handled fairly and in accordance with applicable rules and procedures. This may include advising the member, accompanying them to meetings, and, when necessary, involving legal or other specialist expertise within the union.
When it comes to psychological support, the union advises researchers to contact their employer directly.
At the time of publication, we had not received responses from any of the other Nordic trade unions for academics.
Moving Forward Together
Referring to Dignum’s case, we sought Umeå University’s response to how the case was handled and whether, in retrospect, the university had the structures and support in place to assist a researcher facing an allegation of misconduct, and what lessons it had learned from that case.
In an email, the university confirmed it has procedures and support mechanisms for employees facing difficult situations. These include support related to the work environment and well-being, as well as communication support when requested.
– These procedures and support functions are in place, have been used previously, and have worked well, Lars Nordlander, Human Resources Director at Umeå University, told NordMedia Network.
Nordlander noted that the university does not comment publicly on individual personnel cases. He said Umeå nevertheless continuously evaluates its work and seeks to address any systemic shortcomings that emerge. He added that it was too early to say whether the situation would lead to changes in the university’s procedures.
Dignum, however, questions whether changes to procedures would address what she experienced:
– Procedures are not care. Support “when requested” places the burden on the person who is already isolated. In my case, no one from the university reached out to ask how I was. That is not a procedural shortcoming but a question of leadership, and no procedure will fix it.
When asked to provide advice to researchers who find themselves accused, Dignum says:
– Go to your union or a senior colleague on day one, before the silence sets in; do not wait for the institution to come to you. Do not carry it alone, and do not let it stop the work. And be willing to say plainly that you made ordinary errors, if you did. Pretending otherwise gives accusations of this kind more power than they deserve. And document everything: revision histories, drafts, correspondence. Respond in writing, factually, and do not apologise for things you did not do.