May 14, 2026
Cite me maybe... or get banned
New arXiv policy: 1-year ban for hallucinated references
Fake citations? arXiv is dropping the ban hammer and commenters are cheering
TLDR: arXiv is reportedly cracking down on papers with made-up references by banning offenders for a year, sending a clear message that authors are responsible for what they submit. Commenters mostly love the crackdown, though some are already fighting over where helpful AI ends and punishable nonsense begins.
The academic internet just got a fresh dose of discipline drama. arXiv, the giant online home for research papers, is reportedly bringing in a harsh new rule: if your paper includes made-up references — the kind of fake citations often blamed on careless artificial intelligence tools — you could be hit with a 1-year ban. And if that wasn’t enough of a public shaming vibe, some commenters say repeat access may come with an extra hurdle: future papers must first be accepted by a respected journal or conference.
The reaction? Loud, smug, worried, and a little hilarious. One camp is basically popping champagne. They say science has been drowning in “slop,” and this is the rare internet rule people actually want enforced. The mood from supporters is simple: if your name is on the paper, it’s your job to check it. No excuses, no “the chatbot did it,” no sympathy. One commenter even called posting on arXiv a privilege, not a right, which is about as close as academia gets to reality-TV banishment.
But not everyone is fully relaxed. A nervous undercurrent ran through the discussion: what counts as cheating, and what counts as acceptable AI help? One commenter pushed the debate into sci-fi territory, asking whether a fully machine-generated proof of a famous unsolved math problem would even be allowed. Meanwhile, the most relatable comment came from someone who’d seen a coworker submit a draft with literal AI slop still in the text — and get roasted by reviewers. Translation: yes, people are checking, and yes, they will absolutely find it.
Even the jokes wrote themselves: proofread your paper, or enjoy becoming the main character of Research Twitter.
Key Points
- •Thomas G. Dietterich posted a message to arXiv authors about responsibility for paper contents.
- •The post cites arXiv’s Code of Conduct as stating that each author takes full responsibility for all contents of a paper.
- •The statement says this responsibility applies regardless of how the contents were generated.
- •The article identifies Dietterich as an emeritus professor at Oregon State University and former AAAI president.
- •The article title references a 1-year ban for hallucinated references, but the displayed article content itself does not provide the full policy text or enforcement details.