Some reports have recently gone viral which claimed Google was using your personal emails to train its Gemini AI. However, the company has officially come out on social media, assuring users that this isn’t the case and officially shutting down those rumors. It turns out the confusion stems from older settings that have been around for years, not a sneaky new policy update.
A misunderstanding
If you’ve been scrolling through X (formerly Twitter) or Threads lately, you might have seen some alarming posts warning that Google is secretly using your Gmail data to train its AI models. These posts usually point to a setting in Gmail called “Smart features and personalization” and urge users to turn it off immediately. It caused quite a stir, with even some tech security outlets initially picking up the story before issuing corrections.Google has stepped in to clear the air. In a new post, the company explicitly stated that Gemini is not trained on Gmail content. They clarified that the “Smart features” setting—which powers things like tab sorting and Smart Compose—has been there for a long time and isn’t a new toggle designed to harvest your data for LLMs. The panic seems to be a classic case of internet misinformation spiraling out of control.
Privacy concerns
This situation highlights just how on edge everyone is about AI and data privacy right now. With companies like OpenAI (creators of ChatGPT) facing constant scrutiny over web scraping and data usage, users are understandably hyper-vigilant. It also doesn’t help that the landscape is incredibly competitive. You have Apple pushing its upcoming “Apple Intelligence” as the privacy-focused alternative, largely processing data on-device to avoid these exact kinds of cloud-snooping fears.
For Google, this is a delicate line to walk. While they are heavily in the AI space and need massive amounts of data to compete with GPT-4, they can’t afford to lose the trust of their billions of Workspace users. If people believe their private correspondence is fair game for training a chatbot, they might start looking for alternatives. This dust-up serves as a reminder that tech giants need to be over-communicative about what their “smart” settings actually do, or else the internet will fill in the blanks with the worst-case scenario.
Clearing the air
I’m glad Google responded quickly. It’s easy to see why people panicked—tech companies haven’t exactly earned the benefit of the doubt over the last decade. Personally, I use Gmail’s smart features every day; I rely on that priority inbox to keep my sanity. Knowing that keeping them on doesn’t automatically feed my emails into Gemini is a relief.
That said, I think this is a wake-up call for Google’s UI team. If a legacy setting is vague enough that it can be reframed as a “spyware toggle” in a viral tweet, it might be time to update the description.
Read the latest from Johanna Romero
