It could work with knowledge cut off if they can reliably guarantee it, and also make sure LLMs are not searching github under the surface.
What's the likelihood that the researchers have done this? It seems fairly easy.
What's the likelihood that the researchers have done this? It seems fairly easy.