why would anyone realistically want to do this?
iNaturalist’s built in solution is community review (and various forms of user blocking). the only other way to identify AI generated images is to examine their content credentials to see if they were generated by an AI. however, it’s relatively easy to remove image metadata. so even if there was something in place that blocked images noted as generated by AI in their metadata, it wouldn’t stop a determined person from working around the block.