Social media platforms are increasingly using user-generated content, such as posts, images, and conversations, to train and develop AI systems. Platforms like LinkedIn, X (formerly Twitter), Snapchat, Reddit, and Meta have different policies on how they collect and use this data for AI purposes. Users may have options to opt out on some platforms, but it’s essential to know that if the content is shared publicly, it can often be accessed and used for AI models. Here’s a guide to protecting your data across key platforms.

AI training: How social media posts are being used

OpenAI has acknowledged that creating its AI models like ChatGPT would not have been possible without leveraging copyrighted works. LinkedIn, X, Snapchat, Reddit, and Meta are following suit, utilizing user-generated content to train their artificial intelligence systems. From casual social media posts to selfies, user data is being employed to refine conversational AI, often without explicit user knowledge.

For AI developers, platforms like these are a goldmine of conversational data. Social media posts reflect everyday language and slang, which are essential in making AI more human-like. Additionally, news feeds offer real-time updates, valuable for AI models aiming to keep pace with current events.

Users’ growing concerns over AI use

While AI advancements promise to enhance digital experiences, concerns are mounting among users. Many individuals are uneasy with their personal content, whether a throwaway comment or a private image, being used by multi-billion-dollar corporations to develop AI technologies. David Ogiste, founder of the marketing agency Nobody’s Cafe, shared his thoughts on LinkedIn, telling CNN, “It’s on these platforms to be open about how they will use our data,” adding that he would opt out of using his data for AI training.

Opt-out options on popular platforms

• LinkedIn: Users can choose to opt out of having their data used for AI training. To do so, go to “Settings & Privacy,” select “Data Privacy,” and toggle off “Data for Generative AI Improvement.”
• X (formerly Twitter): Users can opt-out under “Settings,” “Privacy and Safety,” and “Data Sharing and Personalization.” Under the “Grok” section, uncheck the box for AI data use.
• Snapchat: For its “My Selfie” feature, users who opt-in can still disable their face from being used in ads under “Settings” by turning off “See My Selfie in Ads.”
• Reddit: Public posts are shared for AI training, but private content remains untouched. There’s no opt-out for public posts being used for AI purposes.
• Meta: Public posts on Facebook and Instagram are used for AI training, but private messages and content in private accounts are excluded. Users can opt-out by setting their profiles to private.