Summary
The growth in scientific paper submissions to arXiv, which had been steadily rising for decades, showed unusual changes recently. The authors found that after the pandemic, submissions slowed down at two different times. The first slowdown happened before ChatGPT and was mostly due to more AI papers in computer science, while the second slowdown happened after AI tools capable of working on their own were released, but strangely not in AI fields themselves. Surprisingly, submissions outside computer science increased sharply, especially in math, with many more solo-authored papers appearing, suggesting researchers used AI tools that helped them work more independently.
What this means in practice
- •For academic publishers: Adjust editorial and peer review planning by recognizing shifting preprint submission trends driven by autonomous AI tools.
- •For university research offices: Prepare for changing research dissemination patterns by monitoring field-specific shifts in preprint submissions linked to AI-assisted work.
Abstract
As an experiment in agentic assisted analysis, we used an agentic AI model to analyse the effect of agentic AI on arXiv submissions; data acquisition, analysis and the first draft were carried out by the model under the author's direction. Monthly arXiv submissions have grown exponentially for three decades, at about $9\%$ per year. Fitting a piecewise log-linear trend with seasonal modulation to the full record, we find two post-pandemic breaks in growth rate. The first, in mid-2022, precedes ChatGPT and is carried by the expansion of AI research within computer science (CS); outside CS, growth from 2022 to 2025 remained at $8\%$ per year. The second, between December 2025 and February 2026, follows the release of models capable of extended autonomous work and is absent from the AI categories themselves. Since December 2025 submissions without a CS listing have run $17\%$ above trend, and in September 2026 total submissions reached a record $40\,363$. The excess is largest in mathematics, where submissions doubled within a year, and follows a gradient from formal and theoretical fields to experimental ones that is not ordered by surveyed rates of AI-tool adoption. In mathematics the increase is distributed across subfields and authors, two-thirds of it arising from a larger number of distinct authors rather than from a few prolific submitters. Across fields the clearest signature is a shift to single authorship: solo-authored papers rose by $70\%$ within a year in mathematics and condensed matter, and by $44\%$ in quantitative biology even as its total output fell, but not in experimental particle physics. The analysis is deliberately agnostic of LLM-detection methods: it uses only submission counts and metadata, with treatment timing fixed by calendar dates, and is therefore free of the outcome-triggered timing bias of author-level studies.