Bluesky has explained its position on generative Artificial Intelligence (AI) and has assured that it does not train its models with user data nor does it have any “intention to do so” in the future.
The ‘microblogging’ platform has delved into how it uses this technology in its service coinciding with the entry into force of a series of changes related to it and that affect the privacy of users in what is considered its main competitor, X ( formerly Twitter).
Since November 15, X third-party services can “in some cases” use user information “for their own independent purposes,” such as training their AI models.
This will be possible as long as users have activated the permission that allows these services to collect their information, which can be modified in the Settings section and in the ‘Data shared with business partners’ section.
Because “various artists and creators“have recently chosen the social network “as their home” – and this is supported by the new records of Bluesky, which now has around 16 million users – has specified how it trains its AI models to clear doubts from users who have “concerns” about training from other platforms” with your information.
In this sense, he has stressed in a chain of publications of the ‘app’ that he does not use “none” of the content of his subscribers to train the generative AI and that has no “intention to do so” soon.
It has also indicated that it uses this technology “to assist in content moderation”, so that it can classify publications and protect content moderators from content that may be harmful.
Bluesky also uses AI in the Discover algorithmic ‘feed’, to show the publications that it considers interesting to the user, which it also points out in the Terms of its service, where it also points out that uses this data to develop and improve its offers current and future.