Your Gmail might be feeding AI, and even Google’s chatbot is warning about it
With warnings regarding Google’s AI, Gemini, reading your emails, attachments, and bank statements popping up, Gemini itself agrees that tips on how to protect your privacy are worth following and that the tech giant is playing somewhat dirty.
A person looks at Google logo on their phone in Ankara. Dilara Irem Sancar/Anadolu via Getty.
With warnings regarding Google’s AI, Gemini, reading your emails, attachments, and bank statements popping up, Gemini itself agrees that tips on how to protect your privacy are worth following and that the tech giant is playing somewhat dirty.
- Google’s AI assistant Gemini is at the centre of fresh concerns over how much access it has to users’ Gmail and Search data.
- A new default setting called “Search Services History” collects user activity (including media and interactions) to help improve AI systems.
- Privacy advocates say Gemini inside Gmail can process emails to summarise and respond, effectively giving it inbox-level access.
- Google insists it does not train core AI models on personal Gmail content and says email data is not retained.
- Competing services like Proton Mail are using the controversy to urge users to leave Gmail entirely.
The tech giant has started rolling out its new "Search Services History" setting, which is enabled by default for most users globally. This means that it collects data from Google Search services, including images, audio/video recordings, files, and interactions, and uses it to train and improve its generative AI models.
Separately, privacy advocates point out that Gemini processes your emails to summarize threads and answer your prompts, which means, technically, it reads your inbox.
Google itself claims that it does not train its foundational AI models, including Gemini, on users’ personal emails, which remain private, and that Gemini in Gmail does not retain your email data. So basically, in the Gmail case, it comes down to whether you agree that Google's AI reads your inbox to provide you with smart features and whether you trust the company not to abuse this practice.
Therefore, media outlets and competitors are issuing warnings and sharing tips on how users can opt out of this data usage in Google Search and Gmail.
For example, privacy-focused Proton Mail stresses that the Gmail feature "was switched on without consent" and invites Gmail users to switch their email to its service.
Meanwhile, as an “involved party,” Gemini agrees with the suggestion: "If your goal is to stop Gemini from scanning and assisting with your emails in the Gmail app, the steps provided in that link will absolutely do the trick!"
However, the AI chatbot also flags that Proton's warning is "a marketing piece designed to encourage you to leave Gmail and switch to Proton Mail."
A more serious insight from Gemini is that, with its latest privacy-related measures, its developer, Google, is using what privacy advocates call "opt-out fatigue."
"They bundle privacy-invasive AI features into your account by default, banking on the fact that most users are too tired or busy to go digging through their settings to turn them off," Gemini wrote, adding that "taking two minutes to flip this switch is one of the few ways to actively claw back control of your data."
Here's how you can do it:
To stop Gemini from scanning your Gmail
On a Computer (Web Browser):Gmail Website ➔ Gear Icon ➔ See All Settings ➔ General Tab ➔ Scroll to "Smart Features" ➔ UNCHECK "Turn on smart features" ➔ Save Changes.
On Mobile (Gmail App):Gmail App ➔ Three Lines Menu ➔ Settings ➔ Select Your Account ➔ UNCHECK "Smart features and personalization."
To stop Google Search AI media training on your data
On Any Device (Browser):Go to myactivity.google.com ➔ Search Services History ➔ Scroll to Checkboxes ➔ UNCHECK "Save media."
Unlock more exclusive Cybernews content on YouTube.