LinkedIn feels like a diary sometimes — one where people write about bad clients, toxic managers, or HR missteps. The only catch? It’s a diary that 900 million people can read, screenshot, and now, train an AI model on.

Now don’t get me wrong, it’s a great place to celebrate wins and to call out bad experiences. People post screenshots to prove a point or to raise a red flag. That makes total sense. The problem is that those screenshots often contain more than a headline. They can include email addresses, phone numbers, internal ticket numbers, company domains and other details that can be reused by attackers. A single careless screenshot can turn into a map for social engineering, targeted phishing and account compromise.

Beyond the immediate risk of leaking contact details, there is a bigger, newer issue to think about. LinkedIn now plans to use member content to train its generative AI models unless you opt out. That includes posts, comments and other content you create on the platform. If you want to avoid your posts being used in that way, you need to act.

Why this matters

Even if LinkedIn removes direct identifiers before using content for training, the patterns and context in your public posts still feed machine learning. That means private complaints, internal identifiers and other sensitive clues you share publicly can end up contributing to models that learn from real people and real workplaces. Also, the change is being applied by default in many regions, and the opt out window has a hard date attached for some countries. In short, think twice before posting screenshots or sensitive details and check your privacy settings if you care about how your content might be reused.

How to opt out of having your LinkedIn content used for AI training

LinkedIn has added a member setting called Data for Generative AI improvement under the Data privacy section in settings. Turning it off will stop LinkedIn from using your future posts and activity to train content generating AI models. Two important notes before we get into the steps: this setting is on by default for affected regions, and it applies to future data. Content that was already used before you opted out may still be part of training data. If you want to register a stronger objection you can also submit LinkedIn’s Data Processing Objection form.

Follow these steps on desktop. Mobile steps are similar but menu names may be in slightly different places.

1. Log in to LinkedIn.
2. Click your profile picture in the top right and choose Settings & Privacy.
3. In the left column pick Data privacy.
4. Find Data for Generative AI improvement or How LinkedIn uses your data sections.
5. Toggle off Use my data for training content creation AI models or the similarly named option you see there.

If you prefer a shortcut, LinkedIn is exposing a direct preference path where members can change the setting. You can also check LinkedIn’s help center for the current wording and exact menu names because those can change with UI updates.

A few practical tips for safer posting on LinkedIn

• Do not post screenshots that include email headers, ticket numbers, payroll or other internal identifiers. Redact first.
• Blur or crop phone numbers, email addresses and company intranet URLs. A quick edit removes most risk and still lets you share the story.
• Prefer paraphrase over verbatim reproduction. You can still make your point without publishing internal text.
• If you must call out a company, tag the public accounts rather than reposting private documents. That keeps the conversation public and less likely to leak private metadata.
• Periodically audit your older posts and remove or redact anything that could be abused. Old content gets scraped and archived far longer than you expect.

Final word

Posting on LinkedIn is broadcasting. Sometimes that is exactly what you want. Other times it is a mistake waiting to happen. With platforms now using member content to train AI by default in many regions, treating each post as public and permanent is the safest habit. If you are not comfortable with your words being fed into AI models, opt out now and clean up posts that reveal more than they should. Doing that small bit of hygiene protects you and the people you work with.