arXiv Sets One-Year Ban for Unchecked AI-Generated Papers, Terence Tao Backs Direction

AI水论文封一年，署名连坐！arXiv最严新规来了，陶哲轩附议

Thomas Dietterich, chair of arXiv's computer science section, announced a rule that gives all listed authors a one-year ban when a paper contains confirmed unchecked LLM-generated content, and requires post-ban submissions to pass peer review before upload.

Why it matters: HKR-H/K/R all pass: the arXiv rule adds concrete penalties for unchecked LLM content and touches the AI-paper pipeline. This fits 78–84: strong research-ecosystem signal, but not a model or platform launch.