A team of graduate students at Shanghai Jiao Tong University recently asked 14 different AI systems a simple question: given two scientific papers, can you tell which one built on the other? The best AI got it right 27.3% of the time. A coin flip would have done about as well.
You’d think that adding a citation to an AI’s answer would make it more trustworthy. A link to a scientific paper. A reference to a legal precedent. A footnote pointing to a medical journal. Surely the AI has checked its sources, right?
Wrong. A landmark new study accepted at ICML 2026 — one of the world’s most prestigious AI conferences — reveals something deeply counterintuitive: citations make large language models more likely to hallucinate, not less.
Science moves fast. Every day, researchers around the world publish findings that could reshape how we understand the universe, treat diseases, build technology, and live our lives.