Mohammad Hosseini

Scientists have begun using AI agents in tasks such as reviewing the published literature, formulating hypotheses and subjecting them to virtual tests, modeling complex phenomena, and conducting experiments. Although AI agents are likely to enhance the productivity and efficiency of scientific inquiry, their deployment also creates risks for the research enterprise and society, including poor policy decisions based on erroneous, inaccurate, or biased AI works or products; responsibility gaps in scientific research; loss of research jobs, especially entry-level ones; the deskilling of researchers; AI agents’ engagement in unethical research; AI-generated knowledge that is unverifiable by or incomprehensible to humans; and the loss of the insights and courage needed to challenge or critique AI and to engage in whistleblowing. Here, we discuss these risks and argue that, for responsible management of them, reflection on which research tasks should and should not be automated is urgently needed. To ensure responsible use of AI agents in research, institutions should train researchers in AI and algorithmic literacy, bias identification, and output verification, and should encourage understanding of the risks and limitations of AI agents. Research teams may benefit from designating an AI-specific role, such as an AI validator expert or AI guarantor, to oversee and take responsibility for the integrity of AI-assisted contributions.

BackgroundThe emergence of systems based on large language models (LLMs) such as OpenAI’s ChatGPT has created a range of discussions in scholarly circles. Since LLMs generate grammatically correct and mostly relevant (yet sometimes outright wrong, irrelevant or biased) outputs in response to provided prompts, using them in various writing tasks including writing peer review reports could result in improved productivity. Given the significance of peer reviews in the existing scholarly publication landscape, exploring challenges and opportunities of using LLMs in peer review seems urgent. After the generation of the first scholarly outputs with LLMs, we anticipate that peer review reports too would be generated with the help of these systems. However, there are currently no guidelines on how these systems should be used in review tasks.MethodsTo investigate the potential impact of using LLMs on the peer review process, we used five core themes within discussions about peer review suggested by Tennant and Ross-Hellauer. These include 1) reviewers’ role, 2) editors’ role, 3) functions and quality of peer reviews, 4) reproducibility, and 5) the social and epistemic functions of peer reviews. We provide a small-scale exploration of ChatGPT’s performance regarding identified issues.ResultsLLMs have the potential to substantially alter the role of both peer reviewers and editors. Through supporting both actors in efficiently writing constructive reports or decision letters, LLMs can facilitate higher quality review and address issues of review shortage. However, the fundamental opacity of LLMs’ training data, inner workings, data handling, and development processes raise concerns about potential biases, confidentiality and the reproducibility of review reports. Additionally, as editorial work has a prominent function in defining and shaping epistemic communities, as well as negotiating normative frameworks within such communities, partly outsourcing this work to LLMs might have unforeseen consequences for social and epistemic relations within academia. Regarding performance, we identified major enhancements in a short period and expect LLMs to continue developing.ConclusionsWe believe that LLMs are likely to have a profound impact on academia and scholarly communication. While potentially beneficial to the scholarly communication system, many uncertainties remain and their use is not without risks. In particular, concerns about the amplification of existing biases and inequalities in access to appropriate infrastructure warrant further attention. For the moment, we recommend that if LLMs are used to write scholarly reviews and decision letters, reviewers and editors should disclose their use and accept full responsibility for data security and confidentiality, and their reports’ accuracy, tone, reasoning and originality.

Formal recognition of research contributions is critical for career advancement and the allocation of research funding. However, some contributions are mentioned only in the acknowledgments section, which are not indexed by scholarly databases, resulting in little recognition for those involved. We contextualize this shortfall in terms of contributorship, the movement to recognize specific research contributions rather than rely solely on authorship. Broadening the range of recognized individuals is currently advanced largely through reducing authorship restrictions and unbundling manuscripts into smaller elements, such as datasets and protocols, that receive their own attributions. Here we focus on a complementary path, enhancing the contents and metadata of acknowledgments sections. Capitalizing on existing infrastructure and standards, we propose: 1) when acknowledging individuals, authors include ORCIDs (subject to the acknowledgees’ approval) and provide CRediT information where applicable; 2) publishers solicit identities of acknowledgees in a similar way to how they do so for authors in their submission portals; and 3) publishers include metadata of acknowledgees in JATS-XML files. Implementing these steps should encourage scholarly databases to index non-author contributors. The ensuing increase in visibility for research contributors, such as technicians and library professionals, should result in greater recognition of non-author roles.

In a rapidly evolving Generative Artificial Intelligence (GenAI) landscape, researchers, policymakers, and publishers have to continuously redefine responsible research practices. To ensure guidance of GenAI use in research, core principles that remain stable despite technological advancement are needed. This article defines a list of principles guiding the responsible use of GenAI in research, regardless of use case and GenAI technology employed. To define this framework, we conducted an anonymised Delphi consensus procedure comprising a panel of 16 international and multidisciplinary experts in AI, social sciences, law, ethics, and scientific publishing. After three rounds of independent rating and feedback, the panel reached consensus on eight sequentially ordered principles required for responsible GenAI usage: Regulations, Data Security, Quality Control, Originality, Bias Mitigation, Accountability, Transparency, and Broader Impact. For the clear reporting of adherence to these principles, we created a checklist allowing active implementation into the research process. With these efforts, we aim to guide everyday research, support the development of further specified regulations, policies, and guidelines, and promote discussion about GenAI use in research.

Researchers have been using generative artificial intelligence (GenAI) to support writing manuscripts for several years now. However, as GenAI evolves and scientists are using it more frequently, the case for mandatory disclosure of GenAI for writing assistance continues to diverge from the initial justifications for disclosure, namely (1) preventing researchers from taking credit for work done by machines; (2) enabling other researchers to critically evaluate a manuscript and its specific claims; and (3) helping editors determine if a submission satisfies their editorial policies. Our initial position (communicated through previous publications) regarding GenAI use for writing assistance was in favor of mandatory disclosure. Nevertheless, as we show in this paper, we have changed our position and now support instituting a voluntary disclosure policy because currently (1) the credit due to machines for assisting researchers is moving below the threshold of requiring recognition; (2) it is impractical (if not impossible) to accurately specify what parts of the text are human-/GenAI-generated; and (3) disclosures could increase biases against non-native speakers of the English language and compromise the integrity of the peer review system. Consequently, we argue, it should be up to the authors of manuscripts to disclose their use of GenAI for writing assistance. For example, in disciplines where writing is the hallmark of originality, or when authors believe disclosure is beneficial, a voluntary checkbox in manuscript submission systems, visible only after publication (rather than a free-text note in the manuscripts) would be preferable.

Retractions solicited by authors following the discovery of an unintentional error—what we henceforth call a “self-retraction”—are a new phenomenon of growing importance, about which very little is known. Here we present results of a small qualitative study aimed at gaining preliminary insights about circumstances, motivations and beliefs that accompanied the experience of a self-retraction. We identified retraction notes that unambiguously reported an honest error and that had been published between the years 2010 and 2015. We limited our sample to retractions with at least one co-author based in the Netherlands, Belgium, United Kingdom, Germany or a Scandinavian country, and we invited these authors to a semi-structured interview. Fourteen authors accepted our invitation. Contrary to our initial assumptions, most of our interviewees had not originally intended to retract their paper. They had contacted the journal to request a correction and the decision to retract had been made by journal editors. All interviewees reported that having to retract their own publication made them concerned for their scientific reputation and career, often causing considerable stress and anxiety. Interviewees also encountered difficulties in communicating with the journal and recalled other procedural issues that had unnecessarily slowed down the process of self-retraction. Intriguingly, however, all interviewees reported how, contrary to their own expectations, the self-retraction had brought no damage to their reputation and in some cases had actually improved it. We also examined the ethical motivations that interviewees ascribed, retrospectively, to their actions and found that such motivations included a combination of moral and prudential (i.e. pragmatic) considerations. These preliminary results suggest that scientists would welcome innovations to facilitate the process of self-retraction.

Islamic law, as a science, is deeply rooted in rational perception and practical reason, giving it a philosophical structure distinct from modern law. This distinction lies particularly in its metaphysical conception of reason, which ties Islamic law to theology in a metaphysical sense. Unlike modern law, which often reduces legal obligations to political or theological domains, Islamic law bases its framework on a reasoned understanding of “dos and don’ts.” These imperatives are derived from necessity, defined through factual subsistence (thubūt nafs al-amrī) and veritable propositions (qaḍāyā ḥaqīqiyya), rather than simple factual statements (“is and isn’t”). Consequently, Islamic law is considered a realistic science, with its realism stemming from subsistence rather than topical or external determinations. This non-deterministic realism sets it apart from contemporary legal systems, underscoring a fundamental difference in their philosophical foundations.

Traditional conceptions of academic authorship, e.g., the seemingly self-evident assumption that an author is someone who actually writes a text, is challenged by the complexity, scale, and collaborative nature of scientific research. Authors are expected to make a substantial contribution and to assume accountability for all aspects of the work, but in practice, many individuals listed as authors fail to meet all these criteria, notably in biomedical fields. In view of this tension between norm and practice, new conceptions of authorship have emerged, reflecting the growing importance of team science. This paper assesses whether consortium authorship as an emerging practice (also known as ‘group authorship’ or ‘team authorship’) offers a viable approach. Besides practical benefits, there is a normative dimension behind this concept, as it aims to acknowledge the importance of collaboration (seeing it as more than the sum of contributions attributable to individuals), but it also raises ethical questions concerning the responsibilities of consortium authors for the text as a whole. We opt for a case study approach, zooming in on experiences within a research consortium. Besides a literature review, we analyse the results of a deliberative workshop on consortium authorship and analyse how consortium authorship is currently handled in academic journals, notably in the biomedical field. We argue that consortium authorship works best when used in combination with individual authorship, but also notice that it challenges us to rethink the concept of academic authorship as such, for which we use Donna Haraway’s concept of sympoiesis as a starting point.

The book Ethics of Artificial Intelligence offers a solid exploration of arguments and real-world examples that enrich the ongoing debate surrounding AI ethics. With 12 insightful chapters, the book delves into pressing ethical issues, such as the enhancement of human abilities, the nature of consciousness, and questions of responsibility and accountability in various contexts where AI technology is used. This work connects technology ethics with broader philosophical discussions and provides valuable perspectives on the societal implications of AI. Engaging and accessible, it can serve as an essential resource for scholars, technology-enthusiasts, policymakers, and anyone with an interest in the transformative potential of AI and its ethical dimensions.

The ethics of citation has attracted increased attention in recent discussions of research and publication ethics, fraud and plagiarism. Little attempt has been made, however, to situate specific citation misbehaviors in terms of broader ethical practices and principles. To investigate researchers’ perceptions of citation norms, we surveyed active US researchers receiving federal funding from the National Institutes of Health (NIH), the National Science Foundation (NSF), and the National Endowment for the Humanities (NEH). Participants (n = 257) were asked about citation norms they endorse (norm reports), the behaviors they perceive others to engage in (peer reports), and their own citation behaviors (self-reports). Our analyses showed that while considerable discrepancies exist between norm reports, peer reports and self-reports, respondents’ discipline has no significant effect on these. Participants indicated that their own practices and that of their peers falls short of the norms they endorse, but that their own behavior is much less ethically deficient than that of their peers. An exploratory factor analysis revealed that ethically questionable citation behaviors could be grouped usefully into three categories: strategic citations, neglectful citations, and blind citations. Contrary to our hypothesis, the survey showed that greater experience does not always result in better citation practices. A particularly divisive issue pertained to intentionally citing authors from underrepresented demographic groups for reasons of social justice, but broad support for this practice is lacking, although arts and humanities scholars are slightly more supportive. Most researchers view questionable citation practices as negatively affecting their disciplines. Our findings suggest the need for clearer articulations of the citation norms and improved guidance and training about citations.

Journals and publishers are increasingly using artificial intelligence (AI) to screen submissions for potential misconduct, including plagiarism and data or image manipulation. While using AI can enhance the integrity of published manuscripts, it can also increase the risk of false/unsubstantiated allegations. Ambiguities related to journals’ and publishers’ responsibilities concerning fairness and transparency also raise ethical concerns. In this Topic Piece, we offer the following guidance: (1) All cases of suspected misconduct identified by AI tools should be carefully reviewed by humans to verify accuracy and ensure accountability; (2) Journals/publishers that use AI tools to detect misconduct should use only well-tested and reliable tools, remain vigilant concerning forms of misconduct that cannot be detected by these tools, and stay abreast of advancements in technology; (3) Journals/publishers should inform authors about irregularities identified by AI tools and give them a chance to respond before forwarding allegations to their institutions in accordance with Committee on Publication Ethics guidelines; (4) Journals/publishers that use AI tools to detect misconduct should screen all relevant submissions and not just random/purposefully selected submissions; and (5) Journals should inform authors about their definition of misconduct, their use of AI tools to detect misconduct, and their policies and procedures for responding to suspected cases of misconduct.

Generative artificial intelligence (AI) has the potential to transform many aspects of scholarly publishing. Authors, peer reviewers, and editors might use AI in a variety of ways, and those uses might augment their existing work or might instead be intended to replace it. We are editors of bioethics and humanities journals who have been contemplating the implications of this ongoing transformation. We believe that generative AI may pose a threat to the goals that animate our work but could also be valuable for achieving those goals. In the interests of fostering a wider conversation about how generative AI may be used, we have developed a preliminary set of recommendations for its use in scholarly publishing. We hope that the recommendations and rationales set out here will help the scholarly community navigate toward a deeper understanding of the strengths, limits, and challenges of AI for responsible scholarly work.

Generative artificial intelligence (AI) has the potential to transform many aspects of scholarly publishing. Authors, peer reviewers, and editors might use AI in a variety of ways, and those uses might augment their existing work or might instead be intended to replace it. We are editors of bioethics and humanities journals who have been contemplating the implications of this ongoing transformation. We believe that generative AI may pose a threat to the goals that animate our work but could also be valuable for achieving those goals. In the interests of fostering a wider conversation about how generative AI may be used, we have developed a preliminary set of recommendations for its use in scholarly publishing. We hope that the recommendations and rationales set out here will help the scholarly community navigate toward a deeper understanding of the strengths, limits, and challenges of AI for responsible scholarly work.

In this article, we discuss ethical issues related to using and disclosing artificial intelligence (AI) tools, such as ChatGPT and other systems based on large language models (LLMs), to write or edit scholarly manuscripts. Some journals, such as Science, have banned the use of LLMs because of the ethical problems they raise concerning responsible authorship. We argue that this is not a reasonable response to the moral conundrums created by the use of LLMs because bans are unenforceable and would encourage undisclosed use of LLMs. Furthermore, LLMs can be useful in writing, reviewing and editing text, and promote equity in science. Others have argued that LLMs should be mentioned in the acknowledgments since they do not meet all the authorship criteria. We argue that naming LLMs as authors or mentioning them in the acknowledgments are both inappropriate forms of recognition because LLMs do not have free will and therefore cannot be held morally or legally responsible for what they do. Tools in general, and software in particular, are usually cited in-text, followed by being mentioned in the references. We provide suggestions to improve APA Style for referencing ChatGPT to specifically indicate the contributor who used LLMs (because interactions are stored on personal user accounts), the used version and model (because the same version could use different language models and generate dissimilar responses, e.g., ChatGPT May 12 Version GPT3.5 or GPT4), and the time of usage (because LLMs evolve fast and generate dissimilar responses over time). We recommend that researchers who use LLMs: (1) disclose their use in the introduction or methods section to transparently describe details such as used prompts and note which parts of the text are affected, (2) use in-text citations and references (to recognize their used applications and improve findability and indexing), and (3) record and submit their relevant interactions with LLMs as supplementary material or appendices.

This paper analyzes the ethics of social science research employing big data. We begin by highlighting the research gap found on the intersection between big data ethics, SSR and research ethics. We then discuss three aspects of big data SSR which make it warrant special attention from a research ethics angle: the interpretative character of both SSR and big data, complexities of anticipating and managing risks in publication and reuse of big data SSR, and the paucity of regulatory oversight and ethical recommendations on protecting individual subjects as well as societies when conducting big data SSR. Against this backdrop, we propose using David Resnik’s research ethics framework to analyze some of the most pressing ethical issues of big data SSR. Focusing on the principles of honesty, carefulness, openness, efficiency, respect for subjects, and social responsibility, we discuss three clusters of ethical issues: those related to methodological biases and personal prejudices, those connected to risks arising from data availability and reuse, and those leading to individual and social harms. Finally, we advance considerations to observe in developing future ethical guidelines about big data SSR.

This article explores the impact of an Increase in the average Number of Authors per Publication on known ethical issues of authorship. For this purpose, the ten most common ethical issues associated with scholarly authorship are used to set up a taxonomy of existing issues and raise awareness among the community to take precautionary measures and adopt best practices to minimize the negative impact of INAP. We confirm that intense international, interdisciplinary and complex collaborations are necessary, and INAP is an expression of this trend. However, perverse incentives aimed to increase institutional and personal publication counts and egregious instances of guest or honorary authorship are problematic. We argue that whether INAP is due to increased complexity and scale of science, perverse incentives or undeserved authorship, it could negatively affect known ethical issues of authorship at some level. In the long run, INAP depreciates the value of authorship status and may disproportionately impact junior researchers and those who contribute to technical and routine tasks. We provide two suggestions that could reduce the long-term impact of INAP on the reward system of science. First, we suggest further refinement of the CRediT taxonomy including better integration into current systems of attribution and acknowledgement, and better harmony with major authorship guidelines such as those suggested by the ICMJE. Second, we propose adjustments to the academic recognition and promotion systems at an institutional level as well as the introduction of best practices.

BackgroundThe current paper follows up on the results of an exploratory quantitative analysis that compared the publication and citation records of men and women researchers affiliated with the Faculty of Computing and Engineering at Dublin City University (DCU) in Ireland. Quantitative analysis of publications between 2013 and 2018 showed that women researchers had fewer publications, received fewer citations per person, and participated less often in international collaborations. Given the significance of publications for pursuing an academic career, we used qualitative methods to understand these differences and explore factors that, according to women researchers, have contributed to this disparity.MethodsSixteen women researchers from DCU’s Faculty of Computing and Engineering were interviewed using a semi-structured questionnaire. Once interviews were transcribed and anonymised, they were coded by both authors in two rounds using an inductive approach.ResultsInterviewed women believed that their opportunities for research engagement and research funding, collaborations, publications and promotions are negatively impacted by gender roles, implicit gender biases, their own high professional standards, family responsibilities, nationality and negative perceptions of their expertise and accomplishments.ConclusionsOur study has found that women in DCU’s Faculty of Computing and Engineering face challenges that, according to those interviewed, negatively affect their engagement in various research activities, and, therefore, have contributed to their lower publication record. We suggest that while affirmative programmes aiming to correct disparities are necessary, they are more likely to improve organisational culture if they are implemented in parallel with bottom-up initiatives that engage all parties, including men researchers and non-academic partners, to inform and sensitise them about the significance of gender equity.

First introduced in 2015, Contributor Role Ontologies and Taxonomies (CROTs) are the most recent innovation developed to address ethical issues of scholarly authorship. By providing a standard list of roles to specify individual contributions to publications, CROTs enhance transparency and consistency about the reporting of conducted tasks, and accordingly, improve the attribution of credit and responsibilities. Although various parties such as academic journals and universities are increasingly adopting CROTs, thus far no study has evaluated them from an ethical perspective. To fill this gap, this study develops an assessment instrument and subsequently analyses three major CROTs including CRediT, CRO and TaDiRAH, and provides recommendations for improving them from an ethical perspective. This study uses a bottom-up approach comprised of two literature reviews, combined with a review of non-conventional sources (i.e., forums, meeting notes, digital repositories) to collate concerns about authorship and use of CROTs, and generate candidate assessment criteria. Then with a top-down approach and using David Resnik’s 12 principles of research ethics, developed candidate assessment criteria are further analysed, and an assessment instrument consisting of 25 justified criteria is compiled. This instrument is used to identify the ethical strengths and weaknesses of CRediT, CRO and TaDiRAH. Finally, using current best practices and available technology in publication processes, recommendations are provided for further development and improvement of CROTs. Results show that CRediT satisfies more criteria than other CROTs. In relation to the attribution of credit and usability, CRediT provides unique options. However, unlike TaDiRAH that is translated into languages other than English, and CRO that has revised its roles, CRediT is only available in English and has not been updated yet. Improving CROTs benefits almost all parties involved in the development, production, or use of academic publications such as researchers, journal editors, librarians, hiring and promotion committees and funders.

BackgroundThe areas of Research Ethics (RE) and Research Integrity (RI) are rapidly evolving. Cases of research misconduct, other transgressions related to RE and RI, and forms of ethically questionable behaviors have been frequently published. The objective of this scoping review was to collect RE and RI cases, analyze their main characteristics, and discuss how these cases are represented in the scientific literature.MethodsThe search included cases involving a violation of, or misbehavior, poor judgment, or detrimental research practice in relation to a normative framework. A search was conducted in PubMed, Web of Science, SCOPUS, JSTOR, Ovid, and Science Direct in March 2018, without language or date restriction. Data relating to the articles and the cases were extracted from case descriptions.ResultsA total of 14,719 records were identified, and 388 items were included in the qualitative synthesis. The papers contained 500 case descriptions. After applying the eligibility criteria, 238 cases were included in the analysis. In the case analysis, fabrication and falsification were the most frequently tagged violations (44.9%). The non-adherence to pertinent laws and regulations, such as lack of informed consent and REC approval, was the second most frequently tagged violation (15.7%), followed by patient safety issues (11.1%) and plagiarism (6.9%). 80.8% of cases were from the Medical and Health Sciences, 11.5% from the Natural Sciences, 4.3% from Social Sciences, 2.1% from Engineering and Technology, and 1.3% from Humanities. Paper retraction was the most prevalent sanction (45.4%), followed by exclusion from funding applications (35.5%).ConclusionsCase descriptions found in academic journals are dominated by discussions regarding prominent cases and are mainly published in the news section of journals. Our results show that there is an overrepresentation of biomedical research cases over other scientific fields compared to its proportion in scientific publications. The cases mostly involve fabrication, falsification, and patient safety issues. This finding could have a significant impact on the academic representation of misbehaviors. The predominance of fabrication and falsification cases might diverge the attention of the academic community from relevant but less visible violations, and from recently emerging forms of misbehaviors.

The article at hand presents the results of a literature review on the ethical issues related to scientific authorship. These issues are understood as questions and/or concerns about obligations, values or virtues in relation to reporting, authorship and publication of research results. For this purpose, the Web of Science core collection was searched for English resources published between 1945 and 2018, and a total of 324 items were analyzed. Based on the review of the documents, ten ethical themes have been identified, some of which entail several ethical issues. Ranked on the basis of their frequency of occurrence these themes are: 1) attribution, 2) violations of the norms of authorship, 3) bias, 4) responsibility and accountability, 5) authorship order, 6) citations and referencing, 7) definition of authorship, 8) publication strategy, 9) originality, and 10) sanctions. In mapping these themes, the current article explores major ethical issue and provides a critical discussion about the application of codes of conduct, various understandings of culture, and contributing factors to unethical behavior.