ICLR 2027 has introduced new submission rules stipulating that any author can only appear in the author list of a maximum of 20 submissions, with papers exceeding the quota receiving a reminder at the abstract submission stage and random rejection if not adjusted by the full paper deadline. The conference explained that submission numbers are growing too fast, putting pressure on the review system, and quotas are needed to maintain fair and informative reviews. ICLR 2026 received 19,525 valid submissions, with 13,763 papers obtaining formal decisions, involving 76,139 review comments from 18,054 reviewers, compared to about 11,603 submissions for ICLR 2025, an increase of about 68%. Dan Roy, visiting researcher at Google DeepMind and professor at the University of Toronto, Canada, launched a petition to cancel the limit, posting on X that the rule is "ridiculous" and will slow down artificial intelligence development, calling 20 a "completely arbitrary number" and arguing that AI Agents can complete incremental research with little guidance, so everyone should be encouraged to submit as many papers as possible. His post is ironic, reversing ICLR's statements about AI Agents not independently producing top-conference papers and the need to reduce incremental submissions.