AI helped produce two proofs for the same cryptography problem

Last Thursday, MIT graduate student Seyoon Ragavan and UCSB doctoral student Yao-Ting Lin discovered they had independently worked on the same quantum cryptography proof, thanks to OpenAI’s GPT-5.6 Sol Ultra. Both Ragavan and UCSB professor Prabhanjan Ananth, along with UCLA professor Amit Sahai, submitted papers to arXiv.org, crediting the AI model for their breakthroughs. This convergence highlights AI’s growing role in scientific research, where simultaneous discoveries raise questions about independent discovery and credit. The focus of their work, unclonable encryption, leverages quantum information’s unique properties to prevent perfect copying of encrypted messages. This development underscores AI’s potential to solve longstanding problems, transforming how researchers approach open questions in fields like quantum cryptography. QUESTION: How might the increasing use of AI in scientific research change the way we view originality and credit in academic discoveries?