Some jams have a strict no-AI/no-LLM policy, whereas others have an open policy that allows AI/LLM usage. The problem with the latter group is that many hand-coding enthusiasts only want to compare themselves to others in their category, and don't want to play and review a large number of vide-coded games.
An analogy: It is like having a running race where some of the athletes are on performance-enhancing drugs, and others are natural. They can certain share the race track together, but at the end of the race the natural athletes are not going to want to compare themselves to those on performance-enhancing drugs. They can't have the same meaningful discussions ("how did you train for the race?" ... "I just did more drugs. lol."). Likewise those using AI-generation might have separate discussions ("how many tokens did you use?"... "Did you use special harnesses?").
What I'm requesting is a category that can help filter these games apart, in a similar way that the Platform choice already does. Ideally it would be customizable so that the jam organizers can define their own thresholds (e.g., 100% human-made, 90% human-made, Half vibe-coded, 100% AI), but at a minimum just have two groups based on the current binary (Y/N) gen-AI disclosure selection.