Overview
Google Gemini (formerly Bard, formerly LaMDA, formerly whatever Google was calling it that quarter) is Google’s answer to ChatGPT. It comes integrated into Google Search, Gmail, Google Docs, Google Sheets, Android, and Chrome. If you use any Google product — and you almost certainly do — Gemini is now embedded in your workflow whether you asked for it or not.
The product is capable. Multimodal from launch, it handles text, images, audio, and video. The Ultra model competes with GPT-4. The integration with Google’s ecosystem gives it access to your email, your calendar, your documents, your photos, your location history, and your search history. That’s not just an AI assistant. That’s an AI assistant with a 15-year head start on knowing everything about you.
Google has spent the last two decades building the most comprehensive profile of human behavior ever assembled. Gemini is the tool that operationalizes that profile.
What It Knows About You
This is where it gets uncomfortable. Google already knows:
- Every search you’ve ever made (unless you manually deleted it)
- Every email you’ve sent and received via Gmail
- Every document in your Google Drive
- Everywhere you’ve been (Google Maps location history)
- Every YouTube video you’ve watched
- Every photo you’ve stored in Google Photos
- Every website you’ve visited in Chrome
- Every app on your Android phone
- Your contacts, your calendar, your voice (via Google Assistant recordings)
Gemini can now synthesize all of this. When you ask Gemini a question, it doesn’t just search the web — it searches you. “When is my dentist appointment?” pulls from your email and calendar. “What did I photograph in Paris?” pulls from your Photos. “Summarize that article I was reading” pulls from your Chrome history.
The convenience is extraordinary. The surveillance is total.
Google’s privacy policy for Gemini states that human reviewers may read your conversations. Conversations with Gemini are retained for up to 3 years. Even if you turn off Gemini activity, Google retains data for an unspecified period for “safety and abuse” purposes.
The Real Risks
Privacy is off the charts. No other AI product has access to a data profile this comprehensive. OpenAI knows what you type into ChatGPT. Google knows what you type into everything. Gemini is the synthesis layer that turns 20 years of passive data collection into active data utilization. The company that reads your email now has an AI that understands it.
Autonomy erosion runs deep. Gemini isn’t a tool you open — it’s embedded into products you already use. AI summaries appear in your search results whether you want them or not. Gemini suggests replies in Gmail, generates text in Docs, creates formulas in Sheets. Each integration makes Google’s AI more present and your own thinking less necessary. You didn’t subscribe to an AI service. Google simply started doing your thinking for you.
Lock-in is Google’s core strategy. Gmail. Drive. Photos. Maps. Calendar. Chrome. Android. Each product is free. Each product creates dependency. Each product feeds data to every other product. Leaving Google means leaving your email, your documents, your photos, your maps, your browser, and your phone. Adding Gemini to this stack doesn’t increase lock-in — it weaponizes it. The AI makes every Google product better, which makes leaving even harder.
Bias is measurable and documented. Google’s AI has produced racially inaccurate historical images, refused to generate certain content while freely generating others, and exhibited political leanings in its responses. The February 2024 Gemini image generation controversy — where the model refused to generate images of white people in historical contexts — revealed both bias problems and overcorrection problems. The underlying issue: Google is training AI on a biased internet and then applying crude filters to hide the bias rather than address it.
Search quality is degrading. Google’s AI Overviews — AI-generated summaries at the top of search results — have recommended eating rocks, adding glue to pizza, and other dangerous hallucinations. These appear above actual sources, reducing traffic to the websites that provide accurate information. Google is using AI to disintermediate the web that makes Google useful.
Alternatives
- DuckDuckGo: Private search without data profiling. No AI synthesis of your life history.
- ProtonMail + Proton Drive: Privacy-first alternatives to Gmail and Google Drive. End-to-end encrypted.
- Apple ecosystem: iCloud, Apple Maps, Safari. Apple’s business model is hardware, not advertising. Privacy is better, though you’re trading one walled garden for another.
- Self-hosted alternatives: Nextcloud for files, Thunderbird for email. More effort. More freedom.
Our Verdict
Google Gemini earns a D, and it’s a hard D. The product is powerful, the integration is seamless, and that’s exactly the problem. You can opt out of ChatGPT by not visiting the website. You cannot opt out of Gemini without opting out of Google, and opting out of Google in 2026 requires the kind of sustained technical effort most people can’t manage. Google spent two decades collecting your data. Now it has an AI that can use it. The most dangerous AI product isn’t the one you choose to use — it’s the one you can’t avoid.