Your public posts, profile photo included, may be used to train artificial intelligence. In Europe, Meta, LinkedIn and X all now use public content from their users for AI training, and each gives you a way to say no. The catch: opting out only covers future training. Anything already used stays in the models.
This guide explains how to opt out on each platform, what each programme uses, and what your GDPR rights add. Settings menus change often, so the exact labels may move; the details below come from reporting on each programme, checked in October 2026.
Meta: Facebook and Instagram
Meta started training its AI on public content from EU users on 27 May 2025, after Ireland's Data Protection Commission (Meta's lead EU regulator) reviewed its plans.
- What it uses: public posts and comments from adults on Facebook and Instagram, and people's interactions with Meta AI.
- What it says it excludes: private messages, and content from users under 18.
- How to object: Meta sent users a notification with a link to an objection form, and the form is also in its Privacy Centre. You can object at any time.
Meta hasn't clearly said whether an objection affects content it has already used, so the earlier you object, the more you protect.
From 3 November 2025, LinkedIn began using data from members in the EU, EEA and Switzerland, among other regions, to train AI models, with the setting switched on by default.
- What it uses: profile data, including your name, photo, job history, education, location and skills, as well as your posts, articles and comments.
- How to opt out: go to Settings & Privacy → Data privacy → Data for Generative AI Improvement and turn off the option to use your data for training content creation AI models.
- The limit: opting out stops future training on your data. It doesn't take back data that was already used.
X (Twitter) and Grok
X uses public posts and interactions to train its Grok AI.
- How to opt out: go to Settings and privacy → Privacy and safety → Grok & Third-party Collaborators and uncheck the option that allows your public data and interactions to be used for training.
- EU history: after legal action by Ireland's Data Protection Commission in 2024, X agreed to stop using public posts that EU and EEA users made between May and August 2024 for Grok training. The regulator opened a formal inquiry in April 2025 into whether earlier use was lawful.
- The limit: as with LinkedIn, opting out only affects future training.
Your GDPR right to object
These opt-outs exist partly because of the GDPR. When a company processes your personal data based on its "legitimate interests", as Meta does for AI training, Article 21 of the GDPR gives you the right to object, and the company must stop unless it can show compelling grounds that override your interests.
If a platform ignores your objection, you can complain to your national data protection authority: the Autoriteit Persoonsgegevens in the Netherlands, or the Data Protection Authority (GBA/APD) in Belgium. Our guide to GDPR face data rights explains why facial data gets extra protection.
A quick checklist
- Object to Meta AI training through the form in Meta's Privacy Centre.
- Turn off Data for Generative AI Improvement on LinkedIn.
- Uncheck Grok training in X's privacy settings.
- Review which photos are public on each platform, and make old albums private if you don't need them public.
- Repeat after app updates. Settings can move, and new AI features can come with new switches.
AI training isn't the only way public photos get reused
Platform settings only cover photos on that platform. Photos that were copied to other websites, posted by friends, or published in news articles are outside your control there, and face search engines index public websites too. Our guide to opting out of face search engines covers PimEyes, FaceCheck.ID, Lenso.ai and Protevio, and our guide to protecting your photos online covers prevention more broadly.
Sources: Bitdefender on Meta's EU AI training; LexisNexis: Meta can start on 27 May 2025; Irish DPC statement on Meta AI; Malwarebytes on LinkedIn AI training; Anonyome on Grok training; GDPR Article 21. Checked October 2026.
See which photos of you are already public
Know what's out there before you decide what to lock down.
Start a free search →