Bluesky Considers Game-Changing Proposal for User Control Over AI Data Usage
At the recent SXSW conference in Austin, Bluesky CEO Jay Graber revealed significant developments regarding user consent and data usage for generative AI on the platform. This initiative highlights Bluesky’s commitment to user privacy and choice in a landscape increasingly focused on AI capabilities.
Bluesky’s Approach to Data Consent
During her address at SXSW, Graber emphasized the importance of creating a robust framework for user consent regarding how their data could be utilized for generative AI. This is particularly relevant given that Bluesky is a public social network, which enables others to leverage user content for AI training. A notable instance occurred last year when 404 Media identified a dataset comprising 1 million Bluesky posts available on Hugging Face.
Competition with X and User Migration
Bluesky’s largest competitor, X, has been integrating user posts into its AI chatbot, Grok, via its sister company, xAI. Last fall, X modified its privacy policy, permitting third parties to utilize user content for AI training. This decision, alongside the political landscape surrounding owner Elon Musk, contributed to a significant migration of users from X to Bluesky, which has seen its user base grow to over 32 million in just two years.
Future AI Policies at Bluesky
As the demand for AI training data rises, Bluesky is proactively establishing its AI policies. While the platform does not intend to train its own AI systems using user-generated content, Graber highlighted that user consent will be at the forefront of their strategy.
User Choice and Data Rights
Graber stated, “We really believe in user choice,” explaining that users will have the ability to determine how their content is utilized on Bluesky. This system could mirror how websites manage their visibility to search engines through a robots.txt file. Graber added:
- Users will specify their preferences at both the account and post levels.
- Other companies will be asked to respect these consent settings.
Graber noted, “We’ve been working on it with other people in the space concerned about how AI is affecting how we view our data. I think it’s a positive direction to take.”
Conclusion
The proposed framework is currently available on GitHub and aims to ensure that user preferences are honored in the evolving landscape of AI technology. As Bluesky continues to grow, its emphasis on user consent positions it as a leader in responsible data usage in the social media landscape.
