A recent flurry of online reports and social media posts has raised concerns that Google is quietly rolling out an important change to Gmail’s settings, which allegedly defaults users into allowing the platform to access all private messages and attachments for training its Artificial Intelligence (AI) models. This potential access means private communications could be analyzed to refine models that power tools like Smart Compose and AI-generated replies.
The rationale behind such a move would align with Google’s strategy to integrate its powerful Gemini AI into new Gmail features, aiming to enhance productivity through faster email writing and more efficient inbox management. However, the claim that these settings are being switched on by default, rather than requiring explicit user opt-in, is what has fueled significant user discomfort and privacy concerns.
For users handling sensitive or confidential information, this alleged practice is particularly alarming. While Google consistently promises strong privacy measures, including anonymization and data security, the idea of private correspondence being processed for AI training, even with safeguards, is not reassuring to many who demand absolute control over their personal data.
In response to these viral reports, Google has issued a firm denial, labeling the claims as “misleading.” A spokesperson stated that Google has not changed anyone’s settings and reiterated that Gmail’s Smart Features have existed for years. Crucially, Google explicitly stated: “we do not use your Gmail content for training our Gemini AI model.”
The Smart Features in question, which users must enable to receive personalized service, include useful tools beyond predictive text, such as automated order tracking or adding flight details from an email directly to a calendar. Enabling these features in Google Workspace signifies an agreement to let Google use content and activity to personalize the Workspace experience, but Google insists this does not translate to using emails for broad AI training.
Despite Google’s pushback, some reports from individual users—including one Verge staffer—suggest that users who had previously opted out of some Smart Features were mysteriously opted back in, urging caution. This variance in user experience suggests that users should manually double-check their Smart Features personalization settings to ensure their preferences for data usage are accurately maintained.