The dark art of the default
Google claims it respects your privacy, but its entire Gemini rollout is a masterclass in manipulative design. The company says it doesn’t train AI on your Gmail or Drive files directly. Yet when you ask Gemini to summarize an email, that output containing your private data can and will be used to train future models. Google admits this, but buries the true cost behind the promise of a ‘better assistant.’ The real kicker? The only way to stop your conversations from becoming AI training fodder is to permanently delete your chat history. That isn’t a choice. That’s a hostage negotiation.
The labyrinth of opt out
Finding the Gemini privacy controls is an archaeological dig. There is no link in your main Google Account privacy settings where any sane person would look. It is hidden inside the Gemini app under the nondescript label ‘Activity,’ and even then it doesn’t always appear. If you manage to find it and disable AI training, you lose the ability to ever search your past conversations. But the real cruelty is in Gmail. To turn off Gemini features, you must toggle a switch labeled ‘Smart Features.’ Flip it, and you instantly lose inbox filtering tabs, Smart Compose, and package tracking your inbox becomes a firehose of spam. Google even pops up a nag screen begging you to undo it. This isn’t an oversight. It is a textbook ‘forced action’ dark pattern designed to make privacy so painful you give up. Marie Potel of Fair Patterns called Google’s history of hiding privacy settings a practice ‘made to deter people from using them.’ She is right. The company has billions of users trapped by inertia, and it is banking on you not fighting back.
Source: Arstechnica
