Bluesky, competitor attracted more than three million followers since the US presidential election, does not train generative AI models on user data.

In a post published Friday, Bluesky said: “We are not using any of your content to train generative AI, and we have no intention of doing so.” In a follow-up post, it explained that it uses AI to help with content moderation and in the Discover algorithmic feed, before adding “Neither of these systems are Gen AI systems trained on content users”.

The announcement coincided with major changes to X and an influx of new users to Bluesky. Recently, X has changed the way he block function worked and modified its privacy policy, allowing his LLM Grok to learn about user data. This, coupled with Elon Musk’s vocal support for President Trump, appears to have caused an exodus from X – and many are turning to Bluesky.

Bluesky currently has over 17 million userscompared to 9 million users in September. Users are increasingly wary of companies like X, Meta and Google that use their data to train generative AI models without the option to opt out. Since Bluesky is seen as the less toxic version of X, the company’s explicit statement regarding its user data policies is good news for users who are fed up with their data being exploited.

That said, Bluesky currently doesn’t have any generative AI functionality, so it’s easy for the app to say that it doesn’t train anything. As we know from X’s frequent policy updates, this could all change. Bluesky’s terms of service, which was linked in the post, don’t explicitly mention training AI models, so the downside is that a new clause could easily appear. In other words, never say never, but Bluesky users are safe for now.