Comment on Android's new anti-theft features

Hirom@beehaw.org ⁨1⁩ ⁨month⁩ ago

• GingeyBook@lemm.ee ⁨1⁩ ⁨month⁩ ago

  Which is a fraction of a fraction of Android users.

  I’m all for de Googling if that’s what tickles your fancy, but ask anyone on the street and they’ll have no idea what you’re talking about

  source

  • Hirom@beehaw.org ⁨1⁩ ⁨month⁩ ago

    In that blog post, google does not commit to open sourcing these play services features, to integrate in future system upgrade.

    I would love to be proven wrong.

    source

    • FalseMyrmidon@kbin.run ⁨1⁩ ⁨month⁩ ago

      Yep, no one claimed otherwise.

      source
• ReversalHatchery@beehaw.org ⁨1⁩ ⁨month⁩ ago

  While I agree with you, the first step for user centric Android flavors regarding security is to support relocking the bootloader, with a custom (preferably the user’s own) digital signature. As long as we dont have that, an attacker could flash or just boot a custom bootloader through fastboot that does its own thing.

  However that doesn’t really depend on Android system developers, I think, as the problem arises from the inferiority of almost every phone’s bootloader (chain), and probably that can only be reasonably solved by device manufacturers, because as I understand, bootloaders do a lot of heavily device specific things, so there cant really be a common (primary) bootloader, and making one for each phone is a lot of work that also involves lots of reverse engineering, and maybe the early bootloaders cant even be overwritten on some phones…

  source

  • jarfil@beehaw.org ⁨1⁩ ⁨month⁩ ago

    It’s kind of both Google’s and manufacturers responsibility. Google has made available a Dynamic System Updates feature:

    source.android.com/docs/…/dynamic-system-updates

    developer.android.com/topic/dsu

    …but it requires manufacturer support to allow adding custom keys.

    source

    • ReversalHatchery@beehaw.org ⁨1⁩ ⁨month⁩ ago

      Hmm, this is interesting, it looks like if it was a multiboot solution

      source
• BlastboomStrice@mander.xyz ⁨1⁩ ⁨month⁩ ago

  Lol, who wants these anyways. One more reason not to use google play services. (Though I have google play services on this phone.)

  Automatic AI-powered screen lock for when your phone is snatched.

  Theft Detection Lock is a powerful new feature that uses Google AI to sense if someone snatches your phone from your hand and tries to run, bike or drive away. If a common motion associated with theft is detected, your phone screen quickly locks – which helps keep thieves from easily accessing your data.

  One good thing you could try is use this app:

  Find My Device

  source

  • ReversalHatchery@beehaw.org ⁨1⁩ ⁨month⁩ ago

    Theft Detection Lock is a powerful new feature that uses Google AI to sense if someone snatches your phone from your hand and tries to run, bike or drive away. If a common motion associated with theft is detected, your phone screen quickly locks – which helps keep thieves from easily accessing your data.

    Why would we need AI for that? That just makes the function unpredictable. There must be a real solution to detecting this.

    source

    • jarfil@beehaw.org ⁨1⁩ ⁨month⁩ ago

      The real solution is, a heuristic analysis of the phone’s gyroscope and accelerometer data.

      Marketing calls that, “AI”.

      source

      • TehPers@beehaw.org ⁨1⁩ ⁨month⁩ ago

        Not just marketing, that’s the term it’s always been called. Plug a bunch of parameters into a non-deterministic model and you’ve got an AI, at least by what seems to be the common definition of the term.

        source

        • -> View More Comments
• funn@lemy.lol ⁨1⁩ ⁨month⁩ ago

  That’s my concern too

  source

  • jarfil@beehaw.org ⁨1⁩ ⁨month⁩ ago

    It isn’t clear which of these features use Google servers and which ones don’t. The “Find My Device” definitely does, and has no place in AOSP. If they’re actually using AI to compare phone state with some tracked “habitual” behavior, it may also have no place in AOSP, but who knows.

    source