arXiv enforces year-long bans for papers with fabricated citations

arXiv is cracking down on hallucinated references with a new policy: authors who submit papers containing made-up citations face a one-year posting ban. For indie researchers and bootstrapped ML teams, this signals the preprint platform is getting serious about AI-generated content slipping through peer review—a growing problem as language models become more convincing at inventing plausible-sounding papers and authors.