Bluesky, a social network at the forefront of user choice and control, has unveiled a framework for user consent over how their data is used in generative AI. CEO Jay Graber explained that users will have the ability to specify how they want their content used or not used by other companies.
The need for this framework arises from the demand for AI training data and the existing practice of companies using users’ posts without explicit consent. A study last year revealed that 404 Media had accessed a dataset of over 1 million Bluesky user posts hosted on Hugging Face, highlighting the issue of unauthorized data use.
Bluesky’s competitor X has also made changes to its privacy policy to allow third parties to train AI on users’ posts, further fueling the migration of users to Bluesky. The platform now boasts over 32 million users in just two years, demonstrating the growing demand for a more user-centric approach.
Graber emphasized the importance of user choice, stating that users will be able to control how their content is used and whether it can be repurposed by other companies. This framework is similar to how websites specify whether they want to be scraped by search engines or not, but with the added layer of consent for data use in AI.
The proposal, currently available on GitHub, involves obtaining user consent at the account level or post level, with the aim of creating a widely adopted standard that companies and regulators can respect.
Source: https://finance.yahoo.com/news/bluesky-weighing-proposal-gives-users-205559205.html