5 ms·

Gmail app hack works 92 percent of the time

11 points by bcn 12y ago

michel-slm 12y ago
Not the submitter's fault, but the title is really misleading -- contrast with the first paragraph: "Computer scientists have discovered a method of hacking smartphone apps across Android, iOS and Windows devices that is effective up to 92 percent of the time on six of seven popular apps, including Gmail."
- radiospiel 12y ago
  And then the paper the article links to is Android only.
- cfontes 12y ago
  There should be a way to punish head lines like this.
  - michel-slm 12y ago
    In terms of Wired's own classification of wired-tired-expired, I guess it's clear which bucket this falls into
- bcn 12y ago
  Agreed, After submitting and then going to read the paper (http://www.cs.ucr.edu/~zhiyunq/pub/sec14_android_activity_inference.pdf http://www.cs.ucr.edu/~zhiyunq/pub/sec14_android_activity_in...) it's clear the researchers never attempted this attack on anything other than Android, though they do imply that the attack might be possible in other OS.
valleyer 12y ago
Can iOS apps take control of the camera in the background?
scottoreilly 12y ago
I'm pretty sure Wired completely botched this article. It says that the hack works "across Android, iOS and Windows devices" by monitoring "a newly exposed public side channel, which details the shared memory statistics of other processes." That just isn't possible on iOS unless your phone is jailbroken. Sure enough, the source article is only about android. This kind of "journalism" is so frustrating.
userbinator 12y ago
Our attack requires neither specific vulnerabilities nor privacy-sensitive permissions Access to the camera is not a "privacy-sensitive" permission...? That's news to me. Since the delay to get a camera preview frame is only the initialization time (500-1000 ms), the camera is very likely still pointing at the same object, thus obtaining a similar image. Maybe it depends on how people use cameras, but I find that I often start moving the camera away from the subject too soon after hitting the shutter button, resulting in an unrecognisably blurred image. Holding it still for >500ms after that would be a very rare thing for me to do.
Mohobbot 12y ago
http://www.autolike.ga http://www.autolike.ga

11 points by bcn 12y ago

Not the submitter's fault, but the title is really misleading -- contrast with the first paragraph: "Computer scientists have discovered a method of hacking smartphone apps across Android, iOS and Windows devices that is effective up to 92 percent of the time on six of seven popular apps, including Gmail."

radiospiel 12y ago
And then the paper the article links to is Android only.
cfontes 12y ago
There should be a way to punish head lines like this.
- michel-slm 12y ago
  In terms of Wired's own classification of wired-tired-expired, I guess it's clear which bucket this falls into
bcn 12y ago
Agreed, After submitting and then going to read the paper (http://www.cs.ucr.edu/~zhiyunq/pub/sec14_android_activity_inference.pdf http://www.cs.ucr.edu/~zhiyunq/pub/sec14_android_activity_in...) it's clear the researchers never attempted this attack on anything other than Android, though they do imply that the attack might be possible in other OS.

valleyer 12y ago

Can iOS apps take control of the camera in the background?

scottoreilly 12y ago

I'm pretty sure Wired completely botched this article. It says that the hack works "across Android, iOS and Windows devices" by monitoring "a newly exposed public side channel, which details the shared memory statistics of other processes." That just isn't possible on iOS unless your phone is jailbroken. Sure enough, the source article is only about android. This kind of "journalism" is so frustrating.

userbinator 12y ago

Our attack requires neither specific vulnerabilities nor privacy-sensitive permissions Access to the camera is not a "privacy-sensitive" permission...? That's news to me. Since the delay to get a camera preview frame is only the initialization time (500-1000 ms), the camera is very likely still pointing at the same object, thus obtaining a similar image. Maybe it depends on how people use cameras, but I find that I often start moving the camera away from the subject too soon after hitting the shutter button, resulting in an unrecognisably blurred image. Holding it still for >500ms after that would be a very rare thing for me to do.

Mohobbot 12y ago

http://www.autolike.ga http://www.autolike.ga