You’ve probably agreed to let AI companies train on your data — without realizing it. AI models need enormous amounts of training data, and researchers warn they could run out of high-quality public sources. That means your posts, photos, and profile details may already be feeding someone’s model. Here are three hidden consents you can actually take back.
AI Data Privacy (Credits: How-To Geek)
1. Meta AI: Your Facebook and Instagram Posts May Already Be Training Data
Meta uses publicly available data from Instagram and Facebook to train its Meta AI models. That includes public posts and photos — and comments you left under other people’s posts. Worse, this covers content posted long before AI training was ever a thing: if you haven’t touched your Facebook profile in years, your old posts may still have made it into a training set.
The carve-outs: Meta says it does not use the content of private messages unless you actively share them with Meta AI, and posts restricted to Friends or from a private Instagram account are excluded. But note the trap: if your account is private and you comment on a public post, that interaction may still be scraped.
2. LinkedIn: Generative AI Training Is On by Default
LinkedIn has a setting that lets you decide whether your data trains its generative AI models. The catch: the setting is enabled by default. If you never touched it, your name, photos, job history, education, location, and qualifications can all be used for AI training — private messages are excluded.
To opt out, go to Data Privacy and disable Data for Generative AI Improvement. It only stops future use, though — it cannot claw back data already used.
3. X/Grok: Your Tweets Can Train the Bot Even If You Never Used It
X’s Grok chatbot can train on your public posts, profile information, display name, bio, and engagement data — and any conversations you have with Grok itself. You don’t need to be a Grok user for your public data to be in the mix.
Two ways to opt out: make your account private, or go to Privacy and safety > Data sharing and personalization > Grok & Third-party Collaborators and disable the option to let your public data and Grok interactions be used for training.
The bigger lesson: any public post is potential training data now. Tightening privacy settings on the platforms above is the fastest way to limit what AI companies can quietly take from you.
Also Read: Switching to Cloudflare Does Not Hide Your DNS Lookups from Your ISP: One Router Setting Does
Also Read: 6 Pi-hole Mistakes That Can Turn Your Home Network’s Privacy Tool Into a Security Risk
Source: How-To Geek — read the full story there.
