How to configure LinkedIn to not use your data in its AI

¿How to configure LinkedIn so that it does not use your data in its AI? In recent months, LinkedIn has made a significant shift in how it handles its members' information: it has enabled by default the ability to use user data to train artificial intelligence models, both its own and those of partner providers. This decision, according to the platform, seeks to offer more useful features and an improved experience, but it implies that your posts, interactions, and preferences can feed generative algorithms; if you want to prevent LinkedIn from using your data, review your privacy settings.

Although the professional network has been incorporating AI features for some time—from writing assistants to tools that help you better define your application—the contract change has raised concerns. The Microsoft-owned company has reinforced its commitment to ecosystem technology that supports ChatGPT-type systems, suggesting an even closer relationship between LinkedIn data and generative capabilities deployed in their products.

What's changed on LinkedIn and why it affects you

The new policy wording states that LinkedIn and certain vendors may process member information to train models that power generative AI functions. This processing would include content you share, language settings, comments, usage frequency, and activity signals linked to different areas of the service. When the company trains models internally, it claims to apply techniques to reduce identifiable references as far as possible.

In parallel, the platform has expanded its catalog of AI-powered utilities: career-coach-inspired chatbots, resume and cover letter rewriters, and other aids that facilitate everyday tasks for candidates and recruiters. The stated goal is to enhance the match between talent supply and demand and make LinkedIn use more productive, although this entails that part of learning the models rely on community activity.

In several markets, this data usage is enabled without prior explicit consent (opt-out model), meaning you are opted in by default unless you manually disable the options. This approach shifts the burden on the user to review settings and object where appropriate, a sensitive issue for those who emphasize the informed consent and transparency.

Likewise, various communications and updates have introduced temporal nuances: some texts place the implementation of changes in November 2024, and others anticipate expansions of data exchange with Microsoft subsidiaries for purposes of AI and advertising with subsequent entry into force. It is advisable to check the privacy section of your account and how make a LinkedIn private, because option names and scope may vary by region.

Where and who does this policy affect?

LinkedIn has indicated that, as of today, it is not training models with data from residents of the European Union, the European Economic Area, and Switzerland. For the rest of the markets, processing for training purposes may be enabled by default. Recent documents explicitly mention that the use of public content for training purposes in Europe could occur under certain conditions, and that in countries such as the United States or Hong Kong, there would be greater sharing with Microsoft and its affiliates to improve advertising effectiveness.

In any case, the company has implemented a mechanism for users to limit this use. For accounts outside the EU/EEA/Switzerland/UK, a specific switch can be disabled in the settings. For those within those regions, there is a procedure to disable this option. formal objection which is channeled through a form, with follow-up from the Help Center.

Note that even when training is disabled, the company clarifies that some data may be used for other generative AI functions operational within the platform itself (for example, when you interact with a conversational assistant within the platform). This distinction between training models and operational use for specific functions is key to understanding. What exactly does opt-out limit?. Additionally, if you're looking for more control over what's displayed, you can hide content in your feed to reduce exposure.

The way these policies are applied is not static: LinkedIn updates terms and settings screens frequently. Therefore, periodically reviewing the privacy sections will help you detect possible name or scope changes in options such as “Data for Generative AI” or sections linked to advertising and affiliates.

Step by step: How to prevent LinkedIn from using your data to train AI

The most straightforward way is to disable the training permission from your account settings. The route may vary slightly depending on the language and region, but in general terms, the steps are as follows, and they will allow you to limit the use of your information in model training:

1. Log in to your account from the web or app and tap your photo in the top right corner under the menu labeled "Me."
2. Go to “Settings and Privacy” to see all available settings categories.
3. In the side panel, select "Data Privacy" to open the data processing options.
4. Locate the "Data for Generative AI" or "Data to Improve Generative AI" section (the name may vary). Tap and toggle the switch next to "Use my data to train AI models that create content."
5. Save your changes if prompted; you'll see the selector go into a disabled state, reducing the usage of your signals and content in training.

There's another setting you might want to check in certain countries: Under "Settings & Privacy," look for the "Advertising Data" section. There, check if there's an option like "Share data with third parties or affiliates" and leave the toggle off. revoke the exchangeThis helps limit the use of your activity for expanded advertising targeting, including sharing with affiliates.

In addition to the settings above, LinkedIn offers an objection form to object to processing for training purposes. You must complete your first and last name, email address, and a brief explanation of why you do not want the platform to use your personal data for this purpose. After submitting, the system generates a case number that you can check in the Help Center to track the status of your request, although the company warns that there may be delays due to high demandIf you prefer, you can also unsubscribe from LinkedIn.

If you live in the EU, EEA, UK, or Switzerland, the procedure may require this objection route more often than simply using the toggle switch, due to how regional regulations apply. Even so, go to "Data Privacy" and check if the workout setting is listed: if it's visible and active, uncheck it; if it's not, use the opposition form.

What data can be used and where it comes from

LinkedIn's policy covers different types of information. First, there's the data you voluntarily provide: what you include in your profile, the content you publish, the forms you fill out (from surveys to applications), or the documents you attach as an attachment. resume or letter.

There's also information from third parties: people who mention or share details about you in comments, posts, articles, or videos; LinkedIn customers and ecosystem partners; and related entities like Microsoft. This layer of data isn't always under your direct control, but it can influence how you use your data. the systems outline your interests or connections.

Another key source is usage signals: how much time you spend in certain sections, how you interact with posts and ads, what searches you perform, or whether you apply for offers and follow companies. All of this helps models and algorithms infer patterns of activity.

We may add technologies like cookies and similar items, as well as device and location data (e.g., IP address, mobile carrier, or internet provider). This information is used to maintain account security, improve your experience, and potentially feed customization capabilities.

Finally, the communications you make within the network (messages, invitations, events), the data your company or educational institution provides if they purchase LinkedIn services, and the footprint you leave when using third-party services connected to the platform (ads, add-ons, integrations) come into play. When you interact with a generative AI function within LinkedIn, your inputs, the generated results, and the way in which they are processed are all analyzed. you use that tool.

Limitations, legal nuances and what doesn't change when deactivated

An important clarification: disabling the use of your data for training does not erase any learning previously achieved with information that may already be included. In other words, the opt-out acts forward. Furthermore, LinkedIn specifies that this preference does not prevent your data from being used in other generative AI functions operating on the platform itself, for example when you chat with a assistant within LinkedIn.

The underlying debate revolves around consent. The difference between the opt-in model (you enter only if you accept) and the opt-out model (you participate unless you opt out) is substantial. In regions with strict regulations, regulatory pressure has driven more active consent, while in other places, companies have moved toward a system where the user must search and unmark boxes. This asymmetry creates friction and confusion.

Some communications invoke the need to use data to strengthen recruitment products and selection tools, a key front for LinkedIn and Microsoft. There have been cases of large companies using recruitment assistants to reduce selection times, which would explain the demand for real data to achieve competitive levels of accuracy. Without a large and diverse volume, model quality can suffer.

On the user side, there are criticisms about transparency and the opportunity to object. Those who have requested to object through the form have received case numbers and a tracking channel, but the high volume of requests can lead to longer waits than usual. Your best defense is not only to disable what's appropriate, but also to regularly check if new toggles have appeared in the settings.

LinkedIn's communication on the regional scope of training has been explicit on some points (e.g., not training with data from EU/EEA/Swiss residents at certain times), and more open to change on others (e.g., expanding sharing with affiliates for advertising or analytics). Given this patchwork, it's a good idea to adopt a regular review schedule. Data Privacy and Advertising in your account.

A pattern that is repeated throughout the industry

LinkedIn isn't the only case: several services have rewritten their policies to enable the use of user data for AI purposes. Some music platforms have adjusted terms to improve recommendations based on personal signals; large social networks have tried to use public posts in Europe and have encountered organized opposition; providers of conversational assistants are asking for authorization to use conversations and extending retention times; and even storage and transfer services have rectified after criticism for attempting to use shared files as training material.

The common denominator is the hunger for data. Companies see generative AI as a way to create differentiated products, but the balance between that ambition and the user's ability to decide about their information continues to shift. Hence the importance of ensuring that the participation box remains open. “hidden” in plain sight and that there are clear routes to exercising rights.

Best practices for protecting your privacy on LinkedIn

Although the platform offers specific settings, there are habits that add layers of protection. Review "Settings and Privacy" (the "Data Privacy" and "Advertising Data" sections) monthly to confirm that your preferences remain as you left them. Check if new options related to advertising have appeared. training, affiliates or advertisements.

• Reduce the visibility of your public activity (for example, Who can see my profile or your updates), if you don't need that exposure for your professional goals.
• Restrict the use of cookies and similar technologies in the relevant section, where available, to restrict cross-site tracking.
• Before publishing, assess whether the content contains sensitive information (emails, telephone numbers, identifiers) and replace them with non-identifiable data when it is possible.
• Periodically download a copy of your data from the download tool to better understand what the platform stores about your activity.

If you work with AI features within LinkedIn, remember that your input and the way you interact with the tool may be processed to improve that same feature. That doesn't mean they'll automatically be used to train general models if you've opted out, but they may influence the personalized experience what do you get

The reality is that these policies evolve rapidly. Therefore, in addition to adjusting the switches today, it's a good idea to set a reminder in your calendar to repeat this review later. With this routine, you'll be better positioned to maintain control over your policies. your data and your preferences, regardless of how the terms may change in the future.

Taking all of the above into account, the key is to understand the true scope of each adjustment, identify what is limited (model training) and what can remain active (operational AI features), assess regional differences, and use both the "Data for Generative AI" toggle and the objection form and advertising sections; with that approach, you can keep training at bay with your data while still deciding how much you accept customization in your day-to-day life on LinkedIn.