669
Google's Agentic AI wipes user's entire HDD without permission in catastrophic failure
(www.tomshardware.com)
this post was submitted on 03 Dec 2025
669 points (98.1% liked)
Technology
77090 readers
3484 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I think that's the point, the "agent" (whatever that means) is not running in a sandbox.
I imagine the user assumed permissions are small at first, e.g. single directory of the project, but nothing outside of it. That would IMHO be a reasonable model.
They might be wrong about it, clearly, but it doesn't mean they explicitly gave permission.
Edit: they say it in the video, ~7min in, they expected deletion to be scoped within the project directory.
I think the user simply had no idea what they are doing. I read their post and they say they are not a developer anyways, so I guess that explains a lot.
They said in a post: I thought about setting up a virtual machine but didnt want to bother.
I am being a bit hard on them, I assumed they knew what they were doing: Dev, QA, Test, Prod. Code review prior to production etc. But they just grabbed a tool, granted it root to their shell and ran with it.
But they them selves said it caused issues before. And looking at the posts on the antigravity page, lots of people do.
They basically started using a really crappy tool without any supervision as a noob.
He said "I didn't know I needed a seatbelt for AI". LIKE WHAT THE FUCK. Where have you been that you didn't know that these tools make mistakes. You make mistakes. Everything makes mistakes.
If you go to googles antigravity page, I would quick Nope the fuck out. What a shit page.
Edit: 1 more thing: There is a post where one of the users says something along the lines of: "of course I gave the AI full access to my computer, what do I have to hide"? The level of expertise is stupid low....
Edit2: Also, when shown the screen that says "dont allow terminal commands" and also "dont allow auto excution", they decided to turn those off. Also saying well that is tedious.