How to opt out of AI training on social media

Most social media platforms train their AI models on your data by default. If you do not change your settings, your posts, photos, and messages are fair game.

Here is how to turn those settings off.

Why you need to opt out

When you sign up for services like LinkedIn, Meta, or Google, the default agreement allows them to use your account activity to train their generative AI models.

Platform settings

LinkedIn

LinkedIn opts everyone in automatically. To turn it off:

1. Go to Settings & Privacy, then select Data Privacy.
2. Click Data for Generative AI Improvement.
3. Toggle the setting off.

LinkedIn currently requires users to opt out again every 60 days to keep the setting disabled.

Meta (Facebook and Instagram)

Meta trains its models on your public posts, comments, and photos across Facebook and Instagram.

1. Open the Settings menu and go to the Privacy Center.
2. Select AI at Meta.
3. Click View or edit, then Manage.
4. Choose Not allowed to block future training.

This setting only affects future data. Meta does not remove public content they have already scraped.

Google

Google uses your activity across Search, YouTube, and Chrome to train Gemini.

1. Go to your Google Account page and select Data & Privacy.
2. Find the Web & App Activity card.
3. Uncheck Include Chrome history and activity.
4. Turn off Gemini Apps Activity in your account settings.

X (formerly Twitter)

X trains Grok on your posts and interactions by default.

1. Go to Settings, then Privacy and Safety.
2. Select Data sharing and personalization.
3. Turn off Share your data with third parties for AI training.

TikTok

TikTok lets you disable AI training in its mobile app settings.

1. Go to Settings, then Privacy.
2. Select Personalization and Data.
3. Toggle off Use of your data for AI training.

The limits of opting out

Opting out does not undo the past. If a company has already trained a model on your data, they cannot easily extract it. Your opt-out request only stops them from using new posts.

Some platforms also make these menus hard to find, bury them under multiple submenus, or quietly reset your choices during terms-of-service updates.

What you can do next

1. Turn off AI training settings on every platform you use.
2. Limit the visibility of your posts to friends or connections rather than the public.
3. Delete old posts and account history to shrink the data profile companies can scrape.
4. Switch to platforms that do not train models on user content.

How PrivacyGPT helps

Our dashboard tracks AI training policies and opt-out flows across major platforms. You can check the current status of any platform and find direct links to their opt-out pages.

---

Last verified: June 2026.