By James Bentley
Publication Date: 2025-11-24 17:45:00
It’s becoming harder and harder to know what the rules are when it comes to generative AI. With Meta, X, and even the UK government behind opt-out models, it feels like AI is in a “steal first, ask permission later” state. All of that could go some way to explaining why hundreds of thousands of people believed Google was scraping users’ emails to train Google Gemini based on Google’s own wording.
Just last week, a viral X post managed to amass almost 150,000 likes (at the time of writing), claiming, “You have been automatically OPTED IN to allow Gmail to access all your private messages & attachments to train AI models.”
Talking to The Verge, Google spokesperson Jenny Thomson claims, “These 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.”