Concerns Over Privacy in AI Features of Meta, Google, and LinkedIn

Nov 25, 2025
Technology

Photo: Popular social media apps on Apple iPhone including Facebook, Instagram, TikTok, Threads, X (formerly Twitter), and Reddit. (Collected, Al Jazeera, Getty Images)

Staff Report: PNN

As competition intensifies among major tech companies in the United States to develop artificial intelligence (AI) tools, concerns are rising over the extent to which users’ personal information is collected.

Meta, the owner of Facebook, Instagram, Threads, and WhatsApp, as well as Google and LinkedIn, have launched AI features capable of collecting data from public profiles or emails. While Google and LinkedIn allow users to opt out of AI data usage, Meta provides no such option.

Recently, viral posts on social media claimed that these new AI features collect private conversations, photos, and voice messages. Technology analysts warn that opaque policies could spread misinformation and confuse ordinary users.

Meta’s new policy, effective from December 16, will not use private messages, photos, or voice notes directly for AI training. Only public posts, photos, reels, and comments will be used. However, if someone posts publicly under another user’s name, that information may also be included in AI models.

There is no direct way to disable Meta’s AI. Only WhatsApp allows users to turn off AI features per chat.

Viral posts also claimed that Google reads all Gmail emails and attachments for AI purposes. Google denied this, stating that the new AI feature Gemini Deep Research requires separate user permission.

If smart features are enabled, Google can analyze Gmail and Calendar data for email suggestions and other automated functions. Users can disable these smart features in settings to prevent AI from using their data.

Since November 3, LinkedIn has been using U.S. users’ public profiles and posts for AI training, excluding private messages. Microsoft will also use profile data for ad personalization. LinkedIn is the only platform that allows users to explicitly opt out.

There is still no comprehensive federal law in the U.S. regarding tech companies’ use of data, so user protection depends on each platform’s policies. Experts advise users to regularly check privacy settings and read terms of service before using AI features, though this can be tedious.

The extent to which personal data remains secure is unclear, and analysts believe ordinary users’ privacy is most vulnerable amid the AI race.