I chose to not play this one, so I can't offer any feedback.
Regarding attribution/giving credit, given the rules of this jam and the intent behind it, there is a reason that we require participants to make all assets/content themselves unless it is existing Free/Open Culture licenced works. Signal boosting and celebrating the creations of people who have chosen to participate in the Free Software and Open Culture movements is important and valuable. We have disqualified submissions for making use of content that doesn't fit those requirements and/or for not providing proper credit/attribution.
With that in mind, I think it's reasonable to require that the authors and licences of all content that the machine learning synthesis platform you have used be credited as well. I understand that you may not be in a position to do this if you didn't train it yourself, and given that we didn't have explicit rules about AI/machine learning synthesised content, I'm not going to disqualify you here.