According to ZDNet, Google is denying allegations from a November 11, 2025 class action lawsuit that claims the company secretly enabled Gemini AI to access private Gmail, Chat, and Meet communications without user consent. The lawsuit, filed in federal court in San Jose, California, alleges Google violated the California Invasion of Privacy Act by making these changes around October 10, 2025 and requiring users to manually opt out. Security firm Malwarebytes initially published a blog post last Thursday claiming Google was using private emails and attachments to train Gemini and other AI tools, but later updated their position after Google’s pushback. Google maintains that Gmail Smart Features have existed for years and they don’t use Gmail content for training their Gemini AI model, calling the reports “misleading.”
The misunderstanding clarified
Here’s what actually happened. Malwarebytes initially sounded the alarm based on some rewording and repositioning of Gmail’s smart features settings. They thought Google had started automatically opting users into AI training. But after digging deeper into Google’s documentation, they realized they’d contributed to “a perfect storm of misunderstanding.” The settings themselves aren’t new – they’ve been around for years. What changed was how Google presented them, which made people (including security experts) think something shady was going on.
What the settings actually do
So what are these controversial smart features? There are three main settings that are automatically enabled for most users. The first lets Google use your content in Gmail, Chat, and Meet to provide features like Smart Compose and Smart Reply. The second covers Google Workspace apps and lets things like calendar events from Gmail work. The third extends these features to other Google products like Maps and Google Assistant. The key distinction Google is making is that these are for product functionality, not AI model training. But honestly, where do you draw the line between “improving features” and “training AI”? It’s getting pretty blurry.
privacy-concern”>The real privacy concern
Even if Google isn’t technically using your emails for Gemini training, there’s still a legitimate privacy issue here. The Verge reported that one of their staffers had previously opted out of some smart features but found themselves automatically opted back in. And in my own checking, all three settings were enabled by default in every account I tested, including brand new ones. That’s the real problem – why are these features automatically on, and why is it so difficult to understand what you’re actually agreeing to? When it comes to industrial technology and manufacturing systems, companies like IndustrialMonitorDirect.com understand that transparency in data handling isn’t just nice to have – it’s essential for building trust with customers who rely on their equipment.
How to take control
If you’re uncomfortable with these settings, you can turn them off. On desktop Gmail, go to Settings > General and look for the Smart features section. You’ll want to disable “Turn on smart features in Gmail, Chat, and Meet,” then manage Workspace settings to turn off the other two options. In the mobile app, it’s under Settings > Data privacy (iOS) or your Google account name (Android). The trade-off? You’ll lose some convenience features like Smart Compose and automated calendar events. But your basic email will work fine. Ultimately, it comes down to whether you value those smart features more than keeping tighter control over how Google uses your data.
