Realtime AI News
arXiv Caps Submissions at Two Per Month, With Rejections Still Counting
Starting October 1, 2026, arXiv limits each submitter to two paper submissions per calendar month across all categories, and rejected papers still consume the monthly quota. QbitAI reports that the change is meant to ease the review strain caused by a surge in submissions and low-value papers in the AI era.

arXiv, long a symbol of open preprinting, is now metering submissions. From October 1, 2026, each submitter may post at most two papers per calendar month.
The cap applies across fields, with every author and every category treated the same, and it counts submissions rather than accepted papers: even a paper that is eventually rejected uses up one of the month's slots. arXiv also kept a separate limit introduced in 2024, under which no more than three submissions may be active at any time.
The limit applies to the submitter, not to all co-authors. A paper posted by one person costs only that person a slot, and arXiv advises collaborators to agree in advance on who will handle submission.
There is one escape hatch: if a submitter deletes a paper before it is formally announced, the submission does not count against the monthly quota or the active-submission limit. For authors chasing conference deadlines, arXiv's advice is simply to plan submissions earlier.
Official figures explain the tightening. In September 2016 arXiv received 9,869 submissions in a month; by September 2024 that had risen to 20,569; and in September 2026 it reached 40,363. That is more than a fourfold rise over a decade, with the most recent doubling taking only two years.
AI is the fastest-growing part of that surge. QbitAI reports that the cs.AI category under computer science grew more than sixfold between 2024 and 2026, while other categories roughly doubled over the same period. September 2026 alone generated nearly 9,000 support tickets for arXiv staff and moderators.
Beyond volume, arXiv named several quality problems: thin papers, "salami papers" that slice one piece of work into several submissions, and papers that lean heavily on AI writing and are unusually dense. The organization says it does not ban AI-assisted research as long as authors disclose it and the work meets its bar for scholarly value — the problem is that AI has sharply lowered the cost of producing low-value papers.
arXiv frames the change as a temporary speed bump, buying time to reassess what a reasonable publication rate looks like when advanced AI tools enter research, and to upgrade its moderation tooling and process. The underlying tension is that paper generation can scale, while volunteer moderators still have only 24 hours in a day.
Why it matters
The rule changes the rhythm of preprint posting for AI researchers, especially groups that submit in volume. It also puts a harder question on the table: as AI drives writing costs down, can the academic community's review and filtering capacity keep pace?
Nearby Updates
All10/02, 14:43
Yokogawa joins Singapore AI data-centre testbed consortium
Yokogawa has joined a Singapore-based AI data-centre testbed consortium, according to IT Brief Asia. The move highlights how AI compute build-out is converging with data-centre infrastructure, and reflects Singapore's continued push to anchor the region's AI ecosystem.
10/02, 15:01
Pi coding agent reverses course and adds MCP support with 1.0 release
The Pi coding agent hit its 1.0 milestone on Thursday and added Model Context Protocol (MCP) support to its core, reversing creator Mario Zechner's earlier position that MCP was unnecessary. Earendil, the company that acquired Pi in April 2026, said MCP has improved and that its own MCP changes make it easier to integrate other capabilities, including the Jev decision model.
10/02, 15:35
After Meta, Manus Regains Independence and Debuts Manus 2.0 and Cue Agent
Manus has regained its independence after its time under Meta and used the moment to unveil Manus 2.0 and a new AI agent called Cue, according to digitimes. The twin release signals that the company intends to keep shipping in the crowded general-agent market on its own terms.
10/02, 13:20
OpenAI and Synopsys to Build GPT-Synopsys for Chip Design
Synopsys and OpenAI will jointly develop GPT-Synopsys, a specialised model optimised to operate Synopsys EDA tools and run semiconductor design workflows. The deal adds revenue sharing and a global go-to-market partnership, aiming to turn frontier models into native expert users of chip-design and verification tools.