Google is pushing back against viral claims suggesting it secretly trains its Gemini AI on users’ private Gmail messages.
Google Calls Viral Reports “Misleading”
In a statement to The Verge, the company said it “has not changed anyone’s settings,” adding: “Gmail Smart Features have existed for many years, and we do not use your Gmail content for training our Gemini AI model.”
The clarification came after multiple social media posts alleged that Gmail users were “automatically opted in” to allow Google access to their emails and attachments for AI training. Some of these claims were amplified by security firm Malwarebytes, which has since revised its post and described the situation as “a perfect storm of misunderstanding.”
What Google’s Policies Actually Say
Google Workspace’s privacy policy confirms that data typed directly into Gemini — such as prompts inside the app — may be retained and used for model training. However, content inside Workspace apps like Gmail, Docs, and Sheets is not used to train Gemini unless a user explicitly instructs the AI to access that content (for example, asking Gemini to proofread a document).
False Alarms Around Gmail Are Becoming More Common
This marks the second major misinformation wave targeting Gmail in recent months. In September, viral posts falsely claimed that Google had issued a global password-reset warning for all Gmail users.
Concern Over AI Data Use Remains High
Even though the Gmail claims are incorrect, concerns over companies training AI models on user data are growing. Meta and LinkedIn, for instance, have disclosed plans to use some user data for AI training in regions where the law allows.
Google itself recently agreed to pay $1.375 billion to settle allegations that it collected Texans’ biometric data without consent — a reminder of why public suspicion remains high.
Featured image credits: Stephen Phillips via Unsplash
For more stories like it, click the +Follow button at the top of this page to follow us.