ICLR Submission Surge Overwhelms Peer Review

Episode 1 · ICLR 2026 Submissions Top 50,000 as AI-Generated Papers Overwhelm Peer Review (2026-09-19, 12 posts)

Multiple posters report that ICLR 2026 submissions have surpassed roughly 50,000 papers, with the number still climbing a week before the abstract deadline — sparking an outpouring of concern that AI-generated low-quality papers ("AI slop") are overwhelming the peer review system.

Confirmed

• On September 19, @miniapeur posted that ICLR 2026 submissions had exceeded 50,000 and were still growing, lamenting that reviewers face a flood of AI slop and saying they didn't even want to review for this cycle
• OpenAI VP of Research @gklambauer noted the apparent 50,000 submissions, quipping "is this a DDoS attack on machine learning research?"
• @DimitrisPapail said submissions were approaching 45,000, remarking that it's easy to churn out AI-written papers and submit them hoping to get lucky like a coin flip, but what academia actually needs are interesting, cool, slightly crazy papers
• @qberthet confirmed "the submission flood is really coming"

Unconfirmed

• @qberthet cautioned that the current window overlaps NeurIPS decisions and the ICLR abstract deadline, so many unresolved NeurIPS submissions will be resubmitted to ICLR, significantly inflating the total count; the headline number alone doesn't accurately reflect true submission activity, so how representative 50,000 is remains in question
• Reports from various posts range from "approaching 45,000" to "over 50,000," and no official figure has been released

Why it matters

• Exponential growth in submissions far outpaces reviewer capacity; AI-generated content has driven the marginal cost of low-quality submissions toward zero, directly threatening the sustainability of peer review
• If spam submissions and coin-flip tactics become the norm, the value of top-conference acceptance signals and the academic evaluation system itself will be diluted — a key reason several frontline researchers have voiced open frustration

• ICLR Submissions Near 45,000 as Researcher Decries AI-Generated Paper Slop — DimitrisPapail · 2026-09-19
• 50000 papers submitted to ICLR, sparking DDoS-on-research comparison — gklambauer · 2026-09-19
• ICLR hits 50K submissions, reviewers face an impossible workload — miniapeur · 2026-09-19
• ICLR submissions top 50K with a week to go, reviewers brace for AI slop wave — miniapeur · 2026-09-19
• NeurIPS papers stuck in limbo are inflating ICLR submission counts — qberthet · 2026-09-19
• ICLR submissions top 50,000, researchers call for per-author caps — HildeKuehne · 2026-09-20
• ICLR Flooded with 50,000 Submissions as Researcher Jokes It's a 'DDoS on ML Research' — Michael_D_Moor · 2026-09-20
• ICLR draws 58k submissions as AI-for-Science papers flood in — menhguin · 2026-09-20
• ICLR hits 50k submissions as researchers warn of the AI slop era — burny_tech · 2026-09-20
• ICLR tops 50K submissions with a week left, reviewers brace for AI slop — burny_tech · 2026-09-20
• ICLR tops 50K submissions as researcher proposes accepting only top 300 papers a year — burny_tech · 2026-09-20
• ICLR submissions top 50,000 as researcher calls for per-author paper caps — burny_tech · 2026-09-20

Episode 2 · ICLR 2027 Submissions Top 62K, Exceeding All Past Years Combined (2026-09-19, 25 posts)

ICLR 2027 abstract submissions reached roughly 62K by the deadline, surpassing the combined total of all previous ICLR editions and setting yet another record for an AI top conference—pushing the tension between review capacity and the flood of papers to its limit.

Confirmed

• Around the September 19–20 abstract deadline, multiple sources reported rapidly climbing numbers: 13 hours before the cutoff, submission IDs were already approaching 51,000 (InvariantnCauchy), and burnytech later relayed that the count hit about 60K an hour before the deadline.
• According to chuangma of Kyoto University, submissions topped 62K+ at the abstract deadline, beating the previous edition's record (as retweeted by sivareddyg).
• Author trawasthiai submitted an abstract 5 minutes before the cutoff and shared submission number 62263, suggesting the total came in under 63K.
• Google researcher gspandana remarked that this will be a real stress test for peer review (and LLM-assisted reviewing).

Unconfirmed

• All figures above come from submission IDs and personal disclosures; no official statistics have been released yet, and post-deadline withdrawals mean the final submission count remains subject to official confirmation.

Why it matters

• With submissions exceeding the sum of all past editions, reviewer workload is severely out of step with submission volume, putting review quality and the conference's credibility to the test.
• Authors like trawasthiai suspect that part of this runaway growth stems from AI mass-produced low-quality papers (AI slop)—a phenomenon now becoming the ML community's collective concern about scholarly output quality in the LLM era.

• ICLR 2027 hits 62K+ submissions, peer review faces its biggest stress test yet — sivareddyg · 2026-09-19
• Author's submission ID 62263 suggests ICLR 2027 total near 63K papers — trawasthi_ai · 2026-09-19
• ICLR 2027 Submissions Top 50,000 as Deadline Nears — Invariant_n_Cauchy · 2026-09-20
• ICLR Submissions Hit 60,000 With an Hour Left Before Deadline — burny_tech · 2026-09-20
• ICLR 2027 submissions top 60k, exceeding all prior ICLRs combined — peter_richtarik · 2026-09-20
• ICLR 2027 Tops 60,000 Submissions as Author Proposes an AI-Reviewed Fast Track — PTenigma · 2026-09-20
• ICLR 2027 Draws 60k+ Submissions, Exceeding All Prior ICLRs Combined — valentina__py · 2026-09-20
• ICLR 2027 tops 60k submissions as NeurIPS rejects co-authored papers over AI-written reviews — PlisSergey · 2026-09-20
• ICLR Flooded With 50,000 Submissions as NeurIPS Reviewer Laments Declining Quality — burny_tech · 2026-09-20
• ICLR 2027 hits 60k+ submissions as Fleuret declares peer review dead — burny_tech · 2026-09-20
• ICLR 2027 submissions top all previous years (2013–2026) combined — denny_zhou · 2026-09-20
• ICLR 2027 reportedly hits 60k submissions — 20% acceptance would mean 12,000 papers — aran_nayebi · 2026-09-20
• ICLR 2027 submissions exceed all previous years (2013–2026) combined — denny_zhou · 2026-09-20
• ICLR hits 60,000+ submissions, triple last year's count, with ~42,300 expected to reach review — mariyaivasileva · 2026-09-20
• After Scanning 60k ICLR Submissions, Dev Compares Academia to Zimbabwe Currency — tokenbender · 2026-09-20
• ICLR draws 60,000+ submissions, but are even 60 worth reading? — ziv_ravid · 2026-09-20
• ICLR hits 60,000+ submissions; researchers banter over how few are worth reading — AvivTamar1 · 2026-09-20
• ICLR 2027 hits 60k abstracts — more than all previous years combined — my_cat_can_code · 2026-09-20
• ICLR 2027 hits 60k abstracts — more than every previous year combined — my_cat_can_code · 2026-09-20
• ICLR 2027 receives more submissions than all previous years combined — MannyKayy · 2026-09-20

5 more related posts →

Episode 3 · ICLR Submissions Exceed 50,000, Researchers Propose Hybrid Human-LLM Review (2026-09-20, 2 posts)

With over 50,000 ICLR submissions requiring an estimated 150 person-years of review effort, researcher Shayan Kiyani proposes a hybrid human-LLM review pipeline with an AI pre-screening round to address the reviewer shortage.

• With 50k+ ICLR submissions, researchers propose hybrid human-LLM peer review orchestration — HamedSHassani · 2026-09-20
• 50,000 ICLR submissions equal 150 person-years of reviews; proposal: AI-only screening round first — DJiafei · 2026-09-20