Anybody can use the data as long as it’s public facing. It’s not because websites like reddit and getty stomp their feet and want us to pay that we have to.

https://en.m.wikipedia.org/wiki/List_of_datasets_for_machine-learning_research

Reddit is already in every llm model. Until the courts say otherwise clearly (and I highly doubt they will), it’s fair game as it should be.