@arcticfoxieanimates (https://scratch.mit.edu/projects/1270286184) pushed me to make a project on the new TOS clause that has been causing a lil chaos so from your local YAB member I want to say DO NOT PANIC. I don't have any actual extra information but I do have experiences with the conversation of Scratch and AI from YAB, so Imma try my best to debunk and calm. Here are the facts: This is not for AI art/generative AI - this has been scrapped and ST has heard and understood all the community's concerns over it and are ensuring that their implementation of AI (if they even do implement it) will be ethical, education and preserve the essence of Scratch Inferences (this is not fact but what I'm inferring from my knowledge from YAB and the articles): The likely use of this need for our data to train AI is for new AI powered features, NOT GENERATIVE. For example, suggesting blocks, help with bug fixing or suggesting relevant tags (these are just examples, not confirmed features) The reason they have only released an approach statement instead of telling us their intentions is that any features are still in the early stages of development. Now on the topic of third party, this does not mean selling data. Involving third parties could simply mean hiring an external company to help create models. The new clause in the TOS is legally required IF Scratch goes ahead with any AI training, this does not mean they necessarily will. Personal opinion: I think the relaying of information could have been much clearer because I get why people are panicking over what they can read. And for things like data collection there should always be an opt out. FURTHER READING: TOS: https://mitscratch.freshdesk.com/en/support/solutions/articles/4000219182-scratch-terms-of-service Scratch's approach: https://medium.com/scratchteam-blog/introducing-creative-ai-at-scratch-bcdbd6c3efc9
