According to AppleInsider, Google has made Gmail AI training opt-out by default rather than opt-in, meaning your emails and attachments are automatically being used to train AI models for features like Smart Compose and AI-generated replies. The company has buried this setting in two different locations that both need to be disabled, as discovered by security firm MalwareBytes. Users in the EU, Japan, Switzerland, and UK are automatically opted out due to regional regulations, but everyone else needs to manually disable the feature. The setting applies account-wide rather than per-device, so changes made once will protect you across all platforms. AppleInsider strongly recommends disabling this feature since it gives Google access to sensitive data including healthcare communications, bank statements, and confidential workplace emails without explicit user consent.
Why this matters
Here’s the thing: this isn’t just about Google improving their AI features. It’s about them deciding for you that your private conversations are fair game for training data. Think about what’s in your email – medical test results, financial documents, sensitive work discussions. Basically everything you wouldn’t want strangers reading. And now Google’s AI is scanning it all unless you specifically tell them to stop.
What really gets me is the opt-out approach. Why not ask permission first? It feels like they’re banking on most people never finding these settings. I mean, how many of your relatives even know this is happening? Probably zero. That’s why AppleInsider suggests helping less tech-savvy friends and family navigate these privacy settings.
The broader pattern
This isn’t an isolated incident. Remember when Meta wanted perpetual access to upload your entire camera roll? Big tech companies are increasingly treating user data as their personal training ground for AI development. They’re basically saying “we’ll use your stuff unless you explicitly tell us not to.”
And here’s the kicker – even if you trust Google with your data today, what happens if their security gets compromised tomorrow? Or if their policies change? Your sensitive emails could end up in places you never intended. It’s one thing to use anonymized data, but scanning personal communications feels like crossing a line.
What to do now
So here’s your action plan: disable both settings immediately. The first controls Gmail, Chat, and Meet training, while the second handles Google Workspace Smart Features. Miss one, and Google can still use some of your data. The good news? It’s account-wide, so you only need to do this once.
Look, I get that AI needs training data. But shouldn’t companies be transparent about when they’re using your personal communications? And shouldn’t they ask first? Your email contains some of your most private information – it deserves more respect than being treated as free AI food by default.
