What happened

According to a proposed class-action suit filed in San Jose, California, Google “secretly” activated Gemini AI for Gmail, Google Chat, and Meet on or around October 10, 2025, without clearly informing users.

The complaint alleges Gemini now has access to “the entire recorded history” of users’ private communications, including emails, attachments, and chat messages, unless users choose to opt out.

While users can turn Gemini off, the lawsuit argues the setting is buried deep in Google’s privacy menus, making it hard for the average user to find.

The plaintiffs claim this violates California’s Invasion of Privacy Act, which prohibits recording private communications without consent.

According to the complaint, Gemini wasn’t just watching recent chats, it actually had retrospective access to historic communications.

Why it happened

The lawsuit suggests Google’s actions reflect a deeper issue with its transparency: rather than making Gemini clearly opt-in, it was turned on by default.

If true, this raises major questions about whether Google’s design of privacy settings is intentionally or negligently opaque, effectively forcing data collection on users who don’t dig into advanced settings.

The complaint points to a tension built into Google’s AI strategy: Gemini is deeply integrated into its core productivity tools (Gmail, Chat, Meet), so automatically enabling it could be very useful for Google’s AI‑data feedback loops, albeit at the cost of user privacy.