X rival Bluesky says it won't train AI on user posts

Bluesky, an uprising alternative to Elon Musk's X (formerly Twitter), has said it will not train its artificial intelligence models on the posts of its users, leading the debate in favour of the ethical use of generative AI. The announcement also contrasts the AI training policies of X and Mark Zuckerberg's Meta, which has openly admitted to feeding its models the data of user posts.

"A number of artists and creators have made their home on Bluesky, and we hear their concerns with other platforms training on their data. We do not use any of your content to train generative AI, and have no intention of doing so," said Bluesky in a post on its platform, which saw a steep surge in its user base, mostly the X users who decided to leave the Musk's platform after Donald Trump won the US Presidential Election.

The decentralised social media platform, backed by former Twitter CEO Jack Dorsey, said it uses AI internally "to assist in content moderation, which helps us triage posts and shield human moderators from harmful content. Bluesky added it uses AI in the algorithms to power the Discover feed. "None of these are Gen AI systems trained on user content," it said in a follow-up post.

However, data posted on Bluesky's platform is up for grabs for AI companies, such as OpenAI and Google. According to The Verge, the decentralised platform's robot.txt, its policy that informs third parties what to scrape from its website, does not stop the Gen AI companies from extracting the data. "Just as robots.txt files don't always prevent outside companies from crawling those sites, the same applies here," Bluesky spokesperson Emily Liu was quoted as saying in a report by The Verge. “That said, we’d like to do our part to ensure that outside orgs respect user consent and are actively discussing within the team on how to achieve this.”