AAAI-26 Dual Submissions: Novel Challenges

AAAI-26 organizers compared main-track submissions with papers submitted to nine other archival venues whose review periods overlapped, and also searched within the AAAI track. They prioritized pairs using title-and-abstract similarity, applied an LLM-based overlap assessment tool, and manually reviewed the most severe cases. That process led to 141 desk rejections. The paper argues that semantically equivalent submissions rewritten with different wording now outnumber easy-to-detect exact duplicates, and suggests generative AI may be facilitating the trend. Recommendations include policy updates, pre-deadline detection, cross-venue coordination, and adversarial benchmarks.