How to Prevent the Artificial Intelligence You Use from Training on Your Data.

When considering the use of the many generative artificial intelligence bots available in various applications, there are several important aspects to keep in mind. These include ethical issues related to copyright use, the increase in energy demand, and the potential negative impact on human creativity. An additional point that deserves attention is that these tools often learn and adjust their models through human contributions. Although some artificial intelligence companies attempt to anonymize this data, this practice may not be to everyone's liking. However, in most cases, it is possible to disable data collection for training purposes.

It is important to note that disabling AI training is not equivalent to deleting the interaction history with the chatbot, although both actions are related. Conversations can still be used for training before being deleted. Additionally, some users may prefer to keep their chat history but prevent AI companies from using their data for model optimization.

Below are instructions for disabling AI training in various applications that utilize this technology.

ChatGPT

• On the web: Click on your profile picture (top right corner), select Settings > Data Control, and turn off the option "Improve the model for everyone."
• In the mobile app: Tap the menu button (top left corner), then the three dots next to your account name, and look for the Data Control screen, where you can turn off "Improve the model for everyone."

Copilot

• On the web: Click on your account image (top right corner), then on your name and Privacy. You will find two switches to disable: Text model training and Voice model training.
• In the mobile app: The switches are located in a similar place; tap your account image (top right corner), then on Account and Privacy.

Gemini To disable AI training in Gemini, you must turn off chat history. This is done on the web by clicking on Activity (bottom left corner); in the mobile app, tap your account image (top right corner) and then on Gemini app activity. It should be noted that, regardless of the settings, chats will be retained for 72 hours.

Perplexity

• On the web: Click on the gear icon next to your username (bottom right corner), and turn off the AI Data Retention switch (under Account).
• In the mobile app: Tap your account image (top left corner) and look for the AI Data Usage switch.

Grok on X

• On the web: Go to More > Settings and privacy > Privacy and security > Grok and Third-party collaborators and uncheck the box.
• In the mobile app: Tap your account avatar (top left corner) to access the same Settings and privacy menu.

LinkedIn To manage how your information is used on LinkedIn, look for the switch labeled "Data for generative AI improvement."

• On the web: Click on your profile picture at the top of the page, then on Settings and Privacy > Data Privacy.
• In the mobile app: Tap your profile picture (top left corner), then go to Settings > Data Privacy > Data for generative AI improvement.

Meta AI In the case of Meta, especially outside of Europe or the UK, there is a significant limitation on preventing Facebook data from being used for training. In Europe and the UK, users can object to data collection through a specific form, while in the United States, options are more limited. You can also mute Meta's chatbot on Facebook or Instagram using a simplified version of the app.

Other applications For other applications that use AI, it is advisable to explore the settings and privacy policies to understand how data is handled. Policies can vary significantly. Recently, Adobe stated that it will not use user images to train its AI, while Reddit has agreed to train AI models with user posts, leaving little room for action for its users.

Even on platforms that do not use data for AI training, if the content becomes public or is accessible to external developers, it could be captured by AI bots. Therefore, it is wise to be cautious about what is published and shared.