AI in publishing¶
Scientific production in the era of Large Language Models¶
{2601.13187} Scientific production in the era of Large Language Models, Keigo Kusumegi, Xinyu Yang, Paul Ginsparg, Mathijs de Vaan, Toby Stuart, Yian Yin, Science (2025)
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.
Notes
Thanks to all those who publish x2 more papers that I will not have time to read.
Quotes
For peer reviewers and journal editors, this represents a significant issue. As a shortcut to (imperfectly) screen scientific research, writing characteristics are fast becoming uninformative signals just as the quantity of scientific communication surges. As traditional heuristics break down, editors and reviewers may increasingly rely on status markers such as author pedigree and institutional affiliation as signals of quality, ironically counteracting LLMs' democratizing effects on scientific production. One potential response is to leverage the same technology to assist in evaluating manuscripts. Specialized “reviewer agents” could flag methodological inconsistencies, verify claims, and even assess novelty. Whether this scalable approach will help editors and reviewers focus on substance over surface-level signals or introduce new and unforeseen challenges to the scientific process is a critical uncertainty.
Second, in non-LLM-assisted papers across all three repositories, writing complexity is positively associated with manuscript quality as approximated by the probability of publication in a peer-reviewed venue (logistic regressions, Fig. 3E-G). These results confirm prior research showing a positive association between writing complexity and scientific merit (32). Third, and critically, we find a reversal in the relationship between writing complexity and peer-review outcomes for LLM-assisted manuscripts. For these documents, increases in writing complexity are associated with lower peer assessments of scientific merit (Fig. 3E-G).
Publish and Perish: How AI-Accelerated Writing Without Proportional Verification Investment Degrades Scientific Knowledge¶
{2604.05714} Publish and Perish: How AI-Accelerated Writing Without Proportional Verification Investment Degrades Scientific Knowledge, Seok Joon Kwon (2026)
Artificial intelligence tools are accelerating manuscript production far faster than peer review capacity can expand. Applying the theory of constraints from manufacturing science, we formalize this asymmetry through a minimal two-variable ordinary differential equation model coupling review queue evolution and verification quality degradation via an endogenous, queue-pressure-driven review AI adoption mechanism. The causal chain is: writing AI adoption increases submissions, growing the review queue, which drives reviewer AI adoption under pressure, degrading verification quality and reducing net knowledge output. Under empirically informed parameters (writing acceleration {\(\gamma\)} = 2.0, review acceleration {\(\delta\)} = 0.5), the model predicts a deceptive honeymoon where knowledge output peaks at 1.10K0 (circa 2026), followed by paradox onset at t = 6 years (2028) and long-term degradation to 0.68K0 (32% loss), approaching a steady state of 0.60K0 (40% loss). The critical condition for net benefit is {\(\delta\)} > {\(\gamma\)}; the current operating point lies deep in the paradox regime. Empirical validation against NeurIPS, ICLR, arXiv, and bioRxiv submission data shows qualitative consistency with observed post-ChatGPT acceleration patterns. Policy analysis reveals that only combined interventions such as review infrastructure investment paired with institutional quality standards can restore positive knowledge production.
Quotes
Goldratt’s theory of constraints demonstrates that optimizing a non-bottleneck is not merely wasteful but destructive. Our minimal model shows that AI-accelerated writing, unmatched by proportional investment in review infrastructure, produces a paradoxical decrease in verified knowledge output through a clear causal mechanism: queue pressure drives reviewer AI adoption, which degrades verification quality faster than throughput increases.
Notes
From the author:
Hi Daniele, Thanks for reading my paper. Hope all of us can go through this unprecedent hard time for academia. cheers, sjk
LLM hallucinations in the wild: Large-scale evidence from non-existent citations¶
{2605.07723} LLM hallucinations in the wild: Large-scale evidence from non-existent citations, Zhenyue Zhao, Yihe Wang, Toby Stuart, Mathijs De Vaan, Paul Ginsparg, Yian Yin (2026)
Large language models (LLMs) are known to generate plausible but false information across a wide range of contexts, yet the real-world magnitude and consequences of this hallucination problem remain poorly understood. Here we leverage a uniquely verifiable object - scientific citations - to audit 111 million references across 2.5 million papers in arXiv, bioRxiv, SSRN, and PubMed Central. We find a sharp rise in non-existent references following widespread LLM adoption, with a conservative estimate of 146,932 hallucinated citations in 2025 alone. These errors are diffusely embedded across many papers but especially pronounced in fields with rapid AI uptake, in manuscripts with linguistic signatures of AI-assisted writing, and among small and early-career author teams. At the same time, hallucinated references disproportionately assign credit to already prominent and male scholars, suggesting that LLM-generated errors may reinforce existing inequities in scientific recognition. Preprint moderation and journal publication processes capture only a fraction of these errors, suggesting that the spread of hallucinated content has outpaced existing safeguards. Together, these findings demonstrate that LLM hallucinations are infiltrating knowledge production at scale, threatening both the reliability and equity of future scientific discovery as human and AI systems draw on the existing literature.
Quotes
Once published, hallucinated citations diffuse through the bibliographic sources that researchers and AI systems treat as factual. When authors or automated tools attempt to verify a reference, they consult scholarly search engines and citation databases. As non-existent references surface in these systems as standalone bibliographic entries, in addition to appearing in the full text of citing papers, they become part of the permanent record. Cross-validating unmatched references against Google Scholar (SI S2.11), we do indeed find a growing number of entries that cannot be linked to any real publication yet already appear as references in other papers (Fig. 3c). The pattern replicates across all four corpora (Fig. S10). The accumulation of hallucinated citations within bibliometric databases may begin to erode the mechanisms we have to detect them.