Google Denies Gmail Data Used for AI Training Claims

Google has firmly denied viral claims that it is using users' Gmail emails and attachments to train its artificial intelligence models, including the Gemini AI, amid widespread online concerns about privacy settings.

The controversy erupted last week when social media posts and an initial report from cybersecurity firm Malwarebytes suggested that Google had updated its policies to automatically enroll Gmail users in features allowing access to personal data for AI development. These claims spread rapidly on platforms like X, formerly Twitter, prompting users to check and disable what they perceived as new "Smart Features" in their accounts. However, Google quickly responded, stating that no such changes occurred and that Gmail content is not used for training its AI models.

In a post from its official Gmail account on X, Google clarified the situation on Friday. "Let's set the record straight on recent misleading reports," the company wrote. "We have not changed anyone's settings. Gmail Smart Features have existed for many years, and we do not use your Gmail content to train our Gemini AI model." A Google spokesperson echoed this to media outlets, adding, "These reports are...