Unequal Scientific Recognition in the Age of LLMs

Who Gets Cited? Gender- and Majority-Bias in LLM-Driven Reference Selection

Large language models (LLMs) are rapidly being adopted as research assistants, particularly for literature review and reference recommendation, yet little is known about whether they introduce demographic bias into citation workflows. This study systematically investigates gender bias in LLM-driven reference selection using controlled experiments with pseudonymous author names. We evaluate several LLMs (GPT-4o, GPT-4o-mini, Claude Sonnet, and Claude Haiku) by varying gender composition within candidate reference pools and analyzing selection patterns across fields. Our results reveal two forms of bias: a persistent preference for male-authored references and a majority-group bias that favors whichever gender is more prevalent in the candidate pool. These biases are amplified in larger candidate pools and only modestly attenuated by prompt-based mitigation strategies. Field-level analysis indicates that bias magnitude varies across scientific domains, with social sciences showing the least bias. Our findings indicate that LLMs can reinforce or exacerbate existing gender imbalances in scholarly recognition. Effective mitigation strategies are needed to avoid perpetuating existing gender disparities in scientific citation practices before integrating LLMs into high-stakes academic workflows.

Scientific production in the era of Large Language Models

Large Language Models (LLMs) are rapidly reshaping scientific research. We analyze these changes in multiple, large-scale datasets with 2.1M preprints, 28K peer review reports, and 246M online accesses to scientific documents. We find: 1) scientists adopting LLMs to draft manuscripts demonstrate a large increase in paper production, ranging from 23.7-89.3% depending on scientific field and author background, 2) LLM use has reversed the relationship between writing complexity and paper quality, leading to an influx of manuscripts that are linguistically complex but substantively underwhelming, and 3) LLM adopters access and cite more diverse prior work, including books and younger, less-cited documents. These findings highlight a stunning shift in scientific production that will likely require a change in how journals, funding agencies, and tenure committees evaluate scientific works.

Sharing Knowledge Openly: Author Gender, Race/Ethnicity, and Feminist Science

Open science is an increasingly important movement in contemporary social science. Many funders have begun to mandate open practices, newer journals like Socius are open access, and preprint servers have increased readership of paywalled articles. Feminist scholars have written about the values and pitfalls of open research. In this paper, we put those ideas in conversation with each other and empirical data on open-access publishing to develop a feminist framework for thinking about feminist knowledge, data, publishing, and audience. Using a survey of 19,462 social scientists, linked with their entire Web of Science publication history from 2000 to 2023 and all papers posted to SocArXiv, we replicate past work showing that articles by women are less likely to be some types of open access. We extend these findings to scholars of color and feminist scholarship, while also demonstrating a few surprising null results for all three groups.

Faster science, penalties in evaluation, and concerns on quality and impact: Researchers’ use and perceptions of preprints

The preprint ecosystem has expanded rapidly over the past decade, fundamentally altering science communication. Yet, the scholarly community’s attitudes toward this shift remain underexplored. Through a large-scale survey of US and Canadian biomedical scholars, we provide a comprehensive analysis of preprint utilization, perceived impact, and integration into academic credit systems. We find robust engagement across reading, citing, and submitting preprints; however, this activity is driven primarily by a desire for rapid dissemination rather than a foundational commitment to open science. Furthermore, while preprints are valued as networking assets, perceived career penalties during formal academic evaluations stifle broader cultural adoption. Crucially, to navigate the absence of formal peer review, scholars report a heavy reliance on author reputation as a primary heuristic to evaluate a preprint’s credibility and guide their reading and citation decisions. Notably, despite acknowledging preprints’ role in accelerating knowledge sharing, scholars express significant concerns regarding fraud and misinformation, particularly amid declining public trust in science and emerging threats to scientific integrity from artificial intelligence. To resolve these tensions, the preprint ecosystem must evolve beyond prioritizing speed to foster genuine academic dialogue. Simultaneously, evaluation frameworks must adapt to the realities of preprinting, and innovative quality-control mechanisms are urgently needed to balance rapid dissemination with rigorous scientific integrity.

The unintended consequences of large language models as a labor-augmenting technology in science

As a labor-augmenting technology, large language models (LLMs) have the potential to accelerate scientific activity across the research pipeline. But even if LLMs perform on par with human experts at selected tasks, their use will bring unintended consequences as they alter the balance of frictions and inducements that steer the allocation of research effort across projects. Here we develop a simple mathematical model to illustrate. In fields where LLMs are useful primarily as tools for discovering promising projects, researchers will become more selective about what they publish; where they facilitate the process of publishing existing data, researchers will become less selective. By allowing scientists to work more quickly, LLMs raise the opportunity cost of researcher time, creating incentives to refine papers less thoroughly before moving on. Enticing as it is to imagine that, by saving us time on mundane tasks, LLMs will provide us with more time to think deeply and develop projects completely, our results temper such hopes.

The unintended consequences of large language models as a labor-augmenting technology in science

As a labor-augmenting technology, large language models (LLMs) have the potential to accelerate scientific activity across the research pipeline. But even if LLMs perform on par with human experts at selected tasks, their use will bring unintended consequences as they alter the balance of frictions and inducements that steer the allocation of research effort across projects. Here we develop a simple mathematical model to illustrate. In fields where LLMs are useful primarily as tools for discovering promising projects, researchers will become more selective about what they publish; where they facilitate the process of publishing existing data, researchers will become less selective. By allowing scientists to work more quickly, LLMs raise the opportunity cost of researcher time, creating incentives to refine papers less thoroughly before moving on. Enticing as it is to imagine that, by saving us time on mundane tasks, LLMs will provide us with more time to think deeply and develop projects completely, our results temper such hopes.

Topics, Authors, and Institutions in Large Language Model Research: Trends from 17K arXiv Papers

Large language models (LLMs) are dramatically influencing AI research, spurring discussions on what has changed so far and how to shape the field's future. To clarify such questions, we analyze a new dataset of 16,979 LLM-related arXiv papers, focusing on recent trends in 2023 vs. 2018-2022. First, we study disciplinary shifts: LLM research increasingly considers societal impacts, evidenced by 20x growth in LLM submissions to the Computers and Society sub-arXiv. An influx of new authors -- half of all first authors in 2023 -- are entering from non-NLP fields of CS, driving disciplinary expansion. Second, we study industry and academic publishing trends. Surprisingly, industry accounts for a smaller publication share in 2023, largely due to reduced output from Google and other Big Tech companies; universities in Asia are publishing more. Third, we study institutional collaboration: while industry-academic collaborations are common, they tend to focus on the same topics that industry focuses on rather than bridging differences. The most prolific institutions are all US- or China-based, but there is very little cross-country collaboration. We discuss implications around (1) how to support the influx of new authors, (2) how industry trends may affect academics, and (3) possible effects of (the lack of) collaboration.

Topics, Authors, and Institutions in Large Language Model Research: Trends from 17K arXiv Papers

Large language models (LLMs) are dramatically influencing AI research, spurring discussions on what has changed so far and how to shape the field's future. To clarify such questions, we analyze a new dataset of 16,979 LLM-related arXiv papers, focusing on recent trends in 2023 vs. 2018-2022. First, we study disciplinary shifts: LLM research increasingly considers societal impacts, evidenced by 20x growth in LLM submissions to the Computers and Society sub-arXiv. An influx of new authors -- half of all first authors in 2023 -- are entering from non-NLP fields of CS, driving disciplinary expansion. Second, we study industry and academic publishing trends. Surprisingly, industry accounts for a smaller publication share in 2023, largely due to reduced output from Google and other Big Tech companies; universities in Asia are publishing more. Third, we study institutional collaboration: while industry-academic collaborations are common, they tend to focus on the same topics that industry focuses on rather than bridging differences. The most prolific institutions are all US- or China-based, but there is very little cross-country collaboration. We discuss implications around (1) how to support the influx of new authors, (2) how industry trends may affect academics, and (3) possible effects of (the lack of) collaboration.

A call for broadening the altmetrics tent to democratize science outreach

Common altmetrics indices are limited and biased in the social media that they cover. In this Perspective, we highlight how and why altmetrics should broaden its scope to provide more reliable metrics for scientific content and communication.

Citing Less Critically: LLMs Reshape the Rhetoric and Reach of Scientific Citation

Scientific citations carry rhetorical intent. Scholars may cite prior work positively (supporting), negatively (contrasting), or neutrally (mentioning). As large language models (LLMs) increasingly assist scientific writing, whether they reproduce citations with the same rhetorical intent as humans remains unclear. We introduce a masked-citation task to compare human and LLM-generated citation behavior. For each citation context, an LLM generates a replacement citation sentence, producing a counterfactual corpus directly comparable to human citation. We analyze what, whom, and how models cite, using an LLM-as-a-judge to classify citation intent and a 20-million-edge coauthorship network to measure social distance between cited authors. Across six popular LLMs and 1,746 top NLP conference papers (63k+ contexts, 132k+ citations), three patterns emerge: (1) Compared with human citation, LLMs cite significantly less critically; (2) LLMs over-cite popular and older papers, a tendency amplified for contrasting citations where human writing more often draws on recent, niche work; (3) Whereas humans often cite within their close social network, especially for supporting citations, LLMs tend to draw on more socially distant authors. Together, these differences are double-edged: LLM citation reaches beyond a scholar's close collaborators while being less critical and amplifying visibility bias, reshaping the rhetoric and reach of scientific citation.

Scientific Web Claims: A survey of definitions, tasks, datasets and methods

Scientific web claims are seen as scientific claims as observed on the Web, across social media, online news, and other platforms. The growing prevalence of scientific discussions on the Web has intensified the need to process and assess this specific type of claims. Unlike claims from scientific publications, scientific web claims are expressed in lay terms, are often decontextualized, and typically lack proper citations, which poses unique challenges for their identification, verification, and communication. Nevertheless, the correct processing of scientific web claims is crucial to keeping online science discussions accurate and informed, for instance through fact-checking. This survey provides the first systematic overview dedicated specifically to scientific web claims. We review and compare existing definitions, task formulations, datasets, and methodological approaches across three major perspectives: (1) Scientific fact-checking on the Web, (2) Scientific citations on the Web, and (3) Science communication on the Web. Our interdisciplinary analysis integrates insights from natural language processing, information retrieval, artificial intelligence, social sciences, and science communication. We identify major methodological challenges, including the lack of unified definitions, domain-agnostic corpora, and foundational models tailored to science-related online discourse. We also discuss challenges related to the existing interplay between emotions and distortions of science online. By mapping current research efforts and highlighting open problems, this survey lays the groundwork for developing robust datasets, methods, and evaluation frameworks to advance the automated processing of scientific web claims, a necessary capability for strengthening the reliability of science-related online discourse at scale.

LLM use in scholarly writing poses a provenance problem

Nature Machine Intelligence - LLM use in scholarly writing poses a provenance problem

The Discovery Engine: A Framework for AI-Driven Synthesis and Navigation of Scientific Knowledge Landscapes

Scientific progress relies on the effective accumulation, synthesis, and critical evaluation of knowledge. Traditionally, the well-documented, peer reviewed publication served as the primary standard for filtering and disseminating credible findings within the scientific community. Recently, however, we are witnessing an unprecedented acceleration in research output, a veritable explosion of scientific publications across all disciplines [1]. Yet, this very abundance creates a paradox: the sheer volume threatens to overwhelm the mechanisms designed for its assimilation and synthesis. Researchers, even within highly specialized subfields, face an almost insurmountable challenge in keeping abreast of relevant developments, integrating disparate findings, and identifying the truly novel signals amidst the noise [2]. This information overload contributes to disciplinary fragmentation, hindering the cross-pollination of ideas essential for disruptive innovation [3]. Furthermore, persistent concerns regarding "reproducibility crisis" [2], predatory journals, inflation of research areas[4], growing retractions and the potential influences of bibliometrics on research direction [5] highlight systemic challenges in validating and prioritizing scientific contributions to fundamental knowledge.

A Novel Kuhnian Ontology for Epistemic Classification of STM Scholarly Articles

Despite rapid gains in scale, research evaluation still relies on opaque, lagging proxies. To serve the scientific community, we pursue transparency: reproducible, auditable epistemic classification useful for funding and policy. Here we formalize KGX3 as a scenario-based model for mapping Kuhnian stages from research papers, prove determinism of the classification pipeline, and define the epistemic manifold that yields paradigm maps. We report validation across recent corpora, operational complexity at global scale, and governance that preserves interpretability while protecting core IP. The system delivers early, actionable signals of drift, crisis, and shift unavailable to citation metrics or citations-anchored NLP. KGX3 is the latest iteration of a deterministic epistemic engine developed since 2019—originating as Soph.io (2019–2020), advanced as iKuhn (2020–2024), and field-tested through Preprint Watch.

Reclaiming scientific publishing: Our duty to make science freely accessible to all

When we (Camille Thomas and Romain Vaucher speaking) entered academia as graduate students in France and Switzerland, we were enthusiastic about the vast amount of research available with a simple click on our university computers. However, we also quickly felt disheartened by the significant amount of research work we couldn’t access when wrapping up our theses from home. Luckily, pirates existed. Empowered by Aaron Swartz’s Guerilla Open Access Manifesto, Alexandra Elbakyan created Sci-Hub in 2011, the greatest leak of scientific knowledge of the century. We felt right in the middle of an Open Access (OA) revolution that would finally make all scientific articles, old and new, accessible to everyone. Fifteen years later, our hopes as idealistic early-career researchers have been crushed by the oligopolistic model of scientific publishing and the subtle pressures of the “publish or perish” culture that reigns over our career development. In the meantime, publishers like Frontiers, MDPI and Springer Nature, to name a few, have exponentially expanded their number of titles. They have become increasingly exploitative of scholarly manpower, moving far away from the genuine accessibility they allegedly promised under the guise of this OA transformation. Open Access is to be praised, but the way it has been implemented through the mainstream Gold and Green OA models now primarily serves the status quo of large for-profit publishers. These entities hijack public money and voluntary editorial labour for their own profit and that of their shareholders (Butler et al., 2023; Shu and Larivière, 2024). In a nutshell, to offer reader accessibility, the Gold OA model requires authors to pay an Article Processing Charge (APC; around 2,000 $/€, rarely less, and often much more) covered by individual research funds, funding agencies or university library deals. The Green OA model allows authors to upload their accepted, non-formatted manuscripts to repositories after an embargo period. While both models allow compliance with funding agency mandates, true equity and accessibility are ultimately left behind. In an article we recently published (Vaucher and Thomas, 2026), we describe the mechanisms through which we, as researchers, inadvertently contribute to keeping research exclusive while driving the publishing model down an unsustainable path. Just like our broader economy, our publishing model and the ways science is evaluated fuelled a predatory system that demands more papers, funding and prestige at an ever-faster rate (Walter and Mullins, 2019), likely at the expense of quality, diversity and ethics (Frank et al., 2023; Heen and Vogt, 2024). These concerns aren’t entirely new. What is new, however, is the growing realisation among societies, universities and funding bodies that we must move away from this system. Initiatives like the European Diamond Capacity Hub, ALMASI and craft-OA are actively paving a way forward that we, as scientists, have yet to fully embrace. In the geosciences, a collective and concerted effort is currently being made by researchers to provide fairer, more sustainable alternatives through community-driven Diamond OA journals (which feature no APCs and completely free access to published articles). Volcanica (Farquharson and Wadsworth, 2018), Sedimentologika (Thomas et al., 2023), Tektonika (Fernández-Blanco et al., 2023), Seismica (Rowe et al., 2022), Geomorphica (Lefebvre et al., 2025), Open Paleontology (Drage et al., 2024), Advances in Geochemistry and Cosmochemistry (Pourret et al., 2025), Geodynamica, jSEDI and Planetary Research are all recently created, scholarly-run journals funded by university library investments to promote better ways of publishing. Their articles are peer-reviewed, free for readers to access, and free for authors to publish. They rely entirely on the voluntary involvement of scientists running open-source editorial platforms (such as Open Journal Systems), transparent workflows, copyediting, production and final dissemination. Diamond OA journals offer an alternative path for all of us to transform our broken publishing system and reclaim ownership of our own science. These efforts go hand in hand with greater involvement in our academic societies and non-profit publishing initiatives. Ultimately, real transformation can only happen if all of us as researchers realise how inherently unfair and exclusive the current system is to labs and institutions that cannot afford steep Gold OA APCs or paywalled journal subscriptions. It also means we must collectively stop evaluating science based on journal prestige and the flawed metrics they own (Posada and Chen, 2018; Sabel and Larhammar, 2025; Simons, 2008). Only by breaking these habits can we truly make knowledge accessible to all.   References Butler, L.-A., Matthias, L., Simard, M.-A., Mongeon, P., and Haustein, S.: The oligopoly’s shift to open access: How the big five academic publishers profit from article processing charges, Quantitative Science Studies, 4, 778–799, https://doi.org/10.1162/qss_a_00272, 2023. Drage, H. B., Keating, J. N., Nielsen, M. L., Saleh, F., and Hearing, T. W. W.: Open Palaeontology: a new model of diamond open access journal for palaeontology, Open Palaeontology, 1, 1–6, https://doi.org/10.26034/la.opal.2024.6223, 2024. Farquharson, J. I. and Wadsworth, F. B.: Introducing Volcanica: The first diamond open-access journal for volcanology, Volcanica, 1, I–IX, https://doi.org/10.30909/vol.01.01.i-ix, 2018. Fernández-Blanco, D., Lacassin, R., Gouiza, M., Perez-Diaz, L., Magee, C., McCarthy, D., Doré, T., Péron-Pinvidic, G., Kavanagh, J., Bond, C., and Schmitt, R.: Tektonika: The Community-Led Diamond Open-Access Journal for Tectonics and Structural Geology, τeκτoniκa, 1, I–XIII, https://doi.org/10.55575/tektonika2023.1.1.56, 2023. Frank, J., Foster, R., and Pagliari, C.: Open access publishing – noble intention, flawed reality, Social Science & Medicine, 317, 115592, https://doi.org/10.1016/j.socscimed.2022.115592, 2023. Heen, E. and Vogt, H.: Scientific rot: Unsustainable publishing practices threatens trust in medicine, Journal of Evaluation in Clinical Practice, 30, 941–944, https://doi.org/10.1111/jep.13989, 2024. Lefebvre, A., Bosch, R., Burrows, K., Giaime, M., Goodwin, G., Lai, L. S.-H., Stammler, M., and Fernández, R.: Geomorphica: The most accessible journal for the geomorphology community, Geomorphica, 1, https://doi.org/10.59236/geomorphica.v1i1.54, 2025. Posada, A. and Chen, G.: Inequality in Knowledge Production: The Integration of Academic Infrastructure by Big Publishers, in: ELPUB 2018, https://doi.org/10.4000/proceedings.elpub.2018.30, 2018. Pourret, O., Millet, M.-A., Marin-Carbonne, J., Mallik, A., Tierney, J. E., Darling, J. R., Kiseeva, E. S., Torres, M. A., Fonseca, R. O. C., Tartèse, R., Namur, O., Klöcking, M., Matthews, S. W., Dahrén, B., Ickert, R. B., and Board,  the inaugural A. in G. and C. editorial: Equitable Access, Open Science, and the Future of Publishing in Geochemistry and Cosmochemistry, Advances in Geochemistry and Cosmochemistry, 1, https://doi.org/10.33063/agc.v1i1.770, 2025. Rowe, C., Agius, M., Convers, J., Funning, G., Galasso, C., Hicks, S., Huynh, T., Lange, J., Lecocq, T., Mark, H., Okuwaki, R., Ragon, T., Rychert, C., Teplitzky, S., and Van den Ende, M.: The launch of Seismica: a seismic shift in publishing, Seismica, 1, https://doi.org/10.26443/seismica.v1i1.255, 2022. Sabel, B. and Larhammar, D.: Reformation of science publishing: the Stockholm Declaration, R Soc Open Sci., 12, 251805, https://doi.org/10.1098/rsos.251805, 2025. Shu, F. and Larivière, V.: The oligopoly of open access publishing, Scientometrics, 129, 519–536, https://doi.org/10.1007/s11192-023-04876-2, 2024. Simons, K.: The Misused Impact Factor, Science, 322, 165–165, https://doi.org/10.1126/science.1165316, 2008. Thomas, C., Privat, A., Vaucher, R., Spychala, Y., Zuchuat, V., Marchegiano, M., Poyatos-Moré, M., Kane, I., and Chiarella, D.: Sedimentologika: a community-driven diamond open access journal in sedimentology, Sedimentologika, 2023. Vaucher, R. and Thomas, C.: Diamond is the new Green—Why Green Open Access is not a sustainable long-term model for scientific publishing, Sedimentologika, 4, https://doi.org/10.57035/journals/sdk.2026.e41.2397, 2026. Walter, P. and Mullins, D.: From symbiont to parasite: the evolution of for-profit science publishing, Molecular Biology of the Cell, https://doi.org/https://doi.org/10.1091/mbc.E19-03-0147, 2019.  

Do Science <i>Kardashians</i> Get Citation Premium? Self‐Fulfilling Effects of Social Media on Scientific Impact

ABSTRACT We analyze whether the visibility of scientists on social media affects the number of academic citations. We use the global COVID‐19 pandemic as a quasinatural experiment that exogenously increased public attention and the demand for expertise. Using publications on COVID‐related topics by social media stars and their coauthors prior to the outbreak of the pandemic, we find that social media stars' pre‐COVID‐era papers received about – more citations annually per paper after 2019. Quantitatively comparable results are obtained when we use scientists' Kardashian index (K‐index) as a benchmark for stardom, however we find no significant effects when using the intensive margin of scientists' K‐indexes. We provide a brief discussion of policy implications in light of these findings.