New privacy policy on X
Starting November 15, X will begin sharing user data with external companies. This data can be used to train AI models, both generative and others. Users are automatically included unless they choose to opt out of the process. The updated policy clarifies that if users do not opt out, third parties may use their information for commercial purposes.
The global AI race and data
Data from users is becoming an extremely valuable resource in the tech world. Ritesh Kotak, a technology analyst, explains that the more data AI models receive, the more accurate they become. Social media platforms, like X, see data sharing as a lucrative new revenue stream, especially given the intense competition for advertising dollars. Ajay Shrestha, a computer science professor, notes that traditional revenue methods, like advertising, are less effective today.
Licensing agreements and data sales
X is not alone in sharing user data for AI training. For instance, Reddit signed a deal with Google worth $60 million annually. Similarly, Stack Overflow began charging AI companies for access to its data last year. Meta and LinkedIn have also used user data for AI model training, showing how widespread this practice has become in the industry.
Why is user data valuable?
Social media posts provide more than just technical information. According to Kotak, these posts are rich in sentiment and emotion, which is critical for training AI to understand human emotions. This makes user data from platforms like X particularly valuable for building advanced AI models capable of analyzing emotional content.
Can you opt out?
Users can choose to opt out of data sharing, but X has not yet updated its settings to allow users to opt out before the November 15 change. Normally, users can disable data sharing in the privacy settings under the "data sharing with business partners" section. However, Kotak warns that once an AI model is trained using your data, it cannot simply "unlearn" that data, even if you opt out later. This means that once the data is used, it is difficult to reverse the process. These changes reflect how user data is increasingly being monetized, and platforms are becoming more transparent about how they are sharing this valuable resource.
source: CBC