Google Dismisses 'Misleading' Claims About Gmail Content Used for AI Training

Recent viral online discussions suggest that users must disable Gmail’s‘smart features’to prevent their emails from being used to train AI models. However,Googlehas categorically denied these allegations, stating that it does not utilize the content of your emails for artificial intelligence training purposes.

Google is actively refuting widespread social media posts and various articles, includingone by Malwarebytes, which purport that Google has altered its policy to leverage Gmail messages and attachments for training AI models. These claims further assert that the only way to opt out is by deactivating“smart features,”such as spell checking.

Jenny Thomson, a Google spokesperson, informedThe Vergethatthese reports are misleading – we have not changed anyone’s settings, Gmail Smart Features have existed for many years, and we do not use your Gmail content for training our Gemini AI model.

Despite Google’s clarification, users might consider reviewing their settings, as oneVergestaffer reported being opted back into Smart Features after having previously disabled them. InJanuary, Google updated its smart feature personalization settings, enabling users to independently manage these features forGoogle Workspaceand other Google products, such asMapsandWallet.

Beyond basic functions like spell checking, enabling Gmail’s smart features facilitates convenient services such astracking ordersor effortlesslyadding flight details from Gmail to your calendar. While activating this feature in Workspace indicates thatyou agree to let Google Workspace use your Workspace content and activity to personalize your experience across Workspace,according to the settings page, Google explicitly states that this agreementdoes notimply the use of your email content for AI model training.