How to opt out of LinkedIn AI training: the setting, the form, and what's left

The Morse teamPublished November 2, 20268 min read

To opt out of LinkedIn AI training, open Settings & Privacy, choose Data privacy, then Data for Generative AI Improvement, and turn the switch off; the direct link is linkedin.com/mypreferences/d/settings/data-for-ai-improvement. As of October 2026, that one switch stops LinkedIn, and Microsoft where it applies, from training content-generating models on your profile, posts, and AI prompts going forward. It doesn't undo past training, and it doesn't cover your feedback, the AI models that rank your feed, or the data LinkedIn shares with Microsoft for ads. Each of those has its own setting or form, below.

We read LinkedIn's generative AI FAQ, its "Update to our Terms and data use" and "Updates to Settings" help pages, its settings help pages, its Privacy Policy (effective November 3, 2025), its European Regional Privacy Notice (effective August 21, 2026), and its User Agreement on October 24, 2026. Every step and quote links to the LinkedIn page it came from. LinkedIn deserves credit for one thing. It publishes a list of what it trains on and what it doesn't, which many AI companies never do.

How do I turn off LinkedIn AI training on a computer?

Four clicks, using the path LinkedIn's privacy settings help page gives for the Data privacy section:

1. Click the Me icon at the top of your LinkedIn homepage.
2. Select Settings & Privacy from the dropdown.
3. Click Data Privacy in the left rail.
4. Under How LinkedIn uses your data, open Data for Generative AI Improvement and switch it off.

LinkedIn's generative AI FAQ links that setting by name to the direct address above, which opens the same switch.

How do I turn it off in the LinkedIn app on iPhone or Android?

LinkedIn's account settings help page gives one set of app steps for both phones: tap the Me icon (your photo) at the top of the homepage, then tap Settings & Privacy at the bottom of the left pane. From there, tap Data privacy, then Data for Generative AI Improvement, and switch it off. In a phone's web browser, the same page says to tap the Profile icon in the navigation bar, then the Settings & Privacy icon in the upper right.

LinkedIn says you can manage your settings from any device you use to access LinkedIn (LinkedIn Help), so you don't need to repeat this on each phone.

Does LinkedIn use your data to train AI?

Yes, unless you opt out. LinkedIn's FAQ lists six categories it uses "to train our generative AI models that create content":

• Profile data: name, photo, current position, work history, education, location, skills, certifications, endorsements, recommendations, and more.
• GAI usage data: what you type into LinkedIn's AI features, such as prompts, search text, and questions.
• Jobs-related data: screening-question answers and résumés you saved to your account for future use.
• Groups data: "Group activity and group messages."
• Member content: posts, articles, poll responses, contributions, and comments.
• Feedback: support requests, thumbs up or down on AI suggestions, and reports about AI content.

The same page lists what it won't train those models on: private messages "including your InMail and Inbox messages," log-in credentials, payment details, and "member-provided salary data or job application data attributable to a specific member."

Most guides repeat LinkedIn's short version, that training covers your profile and public posts and "does not include your private messages," as the terms update page puts it. The FAQ's full list goes further. Messages you send inside a LinkedIn group are on the training list, and so is anything you type into LinkedIn's AI tools. If you've drafted a cover letter with a LinkedIn AI feature, that draft is GAI usage data.

What doesn't the LinkedIn opt-out cover?

Four things, by LinkedIn's own account:

• Training that already happened. Opting out means LinkedIn and its affiliates "won't use your LinkedIn data to improve models that generate content going forward, but does not affect training that has already taken place" (FAQ).
• Your feedback. Feedback "may be used to improve LinkedIn's generative AI models and not covered by the Data for Generative AI Improvement member setting." The FAQ says LinkedIn may use it "even if you have opted-out."
• Models that don't write content. Models "used to personalize your LinkedIn experience or models used for security, trust, or anti-abuse purposes" sit outside the setting.
• Sharing with Microsoft for ads. That's a separate switch, covered below.

For the second and third, LinkedIn points to its Data Processing Objection form. Its objection help page says the form is for "any other objections to or requests for the restriction of use of personal data" beyond your settings. Say plainly in the form which use you object to: feedback, non-generative model training, or both.

LinkedIn AI training by region, as of October 24, 2026

LinkedIn's rules differ by where you live. This table puts its terms update page, its settings history, its ads data page, and its European notice side by side.

The terms update page is inconsistent with itself. Its summary box names the EU, EEA, Switzerland, Canada, and Hong Kong and leaves out the UK, while its UK section says UK data is used and shared from the same date. Go by the UK section and check your setting.

Guides written in 2024 still say EU and UK members can't be opted in. That changed on November 3, 2025. If you're in the EU and sent LinkedIn's objection form before then, LinkedIn's update page asks you to "check your settings value to make sure your choice to opt out is shown."

How do I stop LinkedIn sharing my data with Microsoft for ads?

Outside the EU, EEA, Switzerland, and the UK, turn off Share data with affiliates and partners under Advertising data in Settings & Privacy. LinkedIn's settings history says that setting, renamed in September 2025 from "Data sharing with trusted partners" in the US, controls sharing with Microsoft "for their ad purposes." Its ads data page adds that to stop Microsoft using LinkedIn data it already received, you go to your Microsoft account's privacy page. In the US, LinkedIn says it automatically opts you out of this sharing if your browser sends a Global Privacy Control signal.

You'll still see Microsoft ads either way. LinkedIn's update page says they "will not be personalized with your LinkedIn data."

Does deleting posts or AI chats help?

It helps going forward, but it doesn't reach a model already trained. LinkedIn's FAQ says what you type into its AI features and what it writes back "will be stored by LinkedIn until you delete it," and points to its data access tool to review and delete. The User Agreement, effective November 3, 2025, gives LinkedIn and its affiliates a "worldwide, transferable and sublicensable" license to what you post. Deleting a post ends that license for it, except where LinkedIn "had already sublicensed others" or other people copied or re-shared it. And the Privacy Policy lets LinkedIn share your "publicly-shared content (such as your public LinkedIn posts)" with affiliates, including Microsoft, "to provide or develop their services," subject to the European notice.

How Morse keeps your drafts out of anyone's training data

Every problem above starts with the same arrangement. The company that hosts what you write also trains on it, and you learn the terms by reading its settings pages. LinkedIn's switch is opt-out, forward-only, and split across three controls and a form. The safest draft is one you never type into LinkedIn's AI tools.

Morse doesn't train AI on your content, sell your data, or show ads, so there's no switch to find and no form to file. Your chat history is stored in your browser, encrypted, and not on Morse's servers (Morse privacy policy). Draft the post, the résumé, or the message to a recruiter in Morse, then paste only the finished text into LinkedIn. What you post publicly is still LinkedIn's to train on under its settings, but the rough drafts, the salary numbers, and the second thoughts stay out of LinkedIn's training data.

• Private, for drafting posts and résumés no host keeps. It's the mode a new chat starts in. Morse has zero data retention agreements with its inference, search, and other providers, so they keep nothing.
• Anonymous, to use GPT, Claude, or Gemini without the provider knowing it's you. Our proxy removes your account, email, and IP address before the message goes out. The provider sees the message and may keep it for a limited time to check for abuse, under its own policy, so leave out names you'd rather not send.
• E2EE, for a job search your current employer must never see. Your browser checks the enclave's hardware attestation, then encrypts the message so only that enclave can read it, and Morse's servers relay text they can't read. It relies on the enclave hardware and attestation working as designed, and comes with Pro, at $19 a month.

Clearing your browser or deleting your account removes your Morse history for good, and Morse can't recover it, so keep a copy of the résumé you settle on. Because Morse doesn't store chats, it usually has very little to give authorities. For how the big chat apps handle training, see which AI companies train on your data, and for Microsoft's own assistant, is Copilot private.

Your checklist for LinkedIn today:

• Training on your profile, posts, and prompts: turn off Data for Generative AI Improvement.
• Feedback and non-generative models: file the Data Processing Objection form and name both.
• Microsoft ads: turn off Share data with affiliates and partners, then check your Microsoft account's ad settings.
• Drafts and AI prompts: write them somewhere that doesn't train AI on your content.

Start a chat free, no card needed, and draft your next post there.