Google's Free AI Comes With a Price

Google’s Gemini now connects to your Gmail, Photos, YouTube history, and Chrome browsing — and as of this week, that’s available to every free U.S. user, not just paying Google One subscribers.

The utility is real. An AI that knows you booked a flight last Tuesday, saw photos from that trip, and watched reviews of local restaurants gives you something a generic chatbot cannot. Context makes answers useful.

But Google’s core business is selling attention to advertisers. They know what you click, what you search, what you buy. What they’ve historically lacked is a clear read on why — the intent behind the behavior. Reading your Gmail and Chrome history hands them that. Not as training data, Google says, but as live context for Gemini’s responses. The distinction matters legally. Whether it matters practically is a different question.

Paid subscribers who opted into Personal Intelligence made a knowing trade. When the default for millions of free users shifts to “AI that reads your email,” that same trade happens at a much larger scale — quietly, as a feature expansion, not a policy change.

It’s opt-in, and Google provides controls to disconnect sources. That’s meaningful. But opt-in features with real utility have a way of becoming the default people never revisit.

The product is useful — just know what you’re trading for it.