Two research teams independently solved one open problem in quantum cryptography using OpenAI's GPT-5.6 Sol Ultra. Their papers were submitted three hours apart, The Decoder reports.
The case raises the question of how to understand independent discovery when researchers use the same model. In one described description, a researcher notes that when an unsolved problem is mentioned, the first thing to check is whether GPT can handle it.
The practical significance of the story lies not only in the problem itself but also in scientific priority: with close results, it becomes important to establish how exactly the model was used and what portion of the solution the researchers performed. The available materials do not clarify this.
editorial commentary
Why it matters
An expected consequence is a tightening of requirements for disclosing the role of AI and for documenting the research process. The next observable signal will be the publication of details about the evidence and its independent verification. Substantial uncertainty remains: the available material contains only metadata and does not allow evaluating the correctness or comparability of the solutions.