Gemini Spark just launched. Tuesday’s I/O event showed off this new AI agent, a 24/7 helper designed to run your life.
We watched it plan a block party.
It handled RSVPs. Chased down ghosters. Built a tracker that updates automatically via Gmail. Smooth? Sure. Convenient? Obviously. Available to testers now, then coming to Google AI Ultra subs and Macs later this week.
But stop for a second.
Ask yourself. Do you actually trust Google with that much stuff?
The privacy hole
Google says the app connections are off by default. They claim it doesn’t read your email indiscriminately.
Okay.
But once you turn those permissions on? Good luck knowing exactly what gets stored or shared.
Google promises it sticks to Google apps. But what about the phone numbers from that block party list? What about the guest emails used for follow-ups that aren’t in G-Suite?
Advanced security is just a promise until it breaks.
Data breaches happen. Always do. It’s not a matter of if, but when.
Screen watching?
Gemini Spark uses Gemini 3.5 Flash. It runs in the background. Even when your laptop is off.
That part is unsettling.
It runs independently under your direction? Sounds like it watches your screen to get things done. I don’t like the idea of an AI scanning my sheets for “important info” while I’m trying to work. Or sleep.
Imagine it ordering snacks from Instacart. It knows your address. Your card info. Your routine.
One glitch. One hack. Too many oranges, sure, but also exposed financials.
Is it worth the convenience? Maybe not. I’ll pass on the invite.
What about you? 🕵️♀️
