Hackers can tell what people are typing just by listening through a smartphone's microphone

jim bell jdb10987 at yahoo.com
Sun Aug 25 14:27:59 PDT 2019


 

    On Sunday, August 25, 2019, 02:05:48 PM PDT, Razer <g2s at riseup.net> wrote:  
On August 24, 2019 11:25:42 PM PDT, jim bell <jdb10987 at yahoo.com> wrote:
>https://www.newsbreakapp.com/n/0MbloCip?s=a3&pd=02h1yVjC
>
>Hackers can tell what people are typing just by listening through a
>smartphone's microphone
>
>Daily Mail
>
>
>Hackers could be able to tell what you are typing on your computer just
>by listening to your keystrokes through your smartphone's microphone, a
>new study warns.
>
>Shared from News Break 
>
>
>>Jim Bell's comment:Peter Wright's 1970's book, Spycatcher, described an
>>early, primitive version of this technology.

>By audio waveform signatures. I always assumed this was a 'thing'.

>But if the audio is ever 'detached' from the source person's ID it becomes sort of useless.

I am curious how this is currently done.  In Peter Wright's day (MI5), the typewriters that would be listened to might have been IBM Selectrics.  In that system, the type ball makes multiple separate motions, rotating and changing angle, before the character is automatically struck.  That should contain a wealth of information.
Similarly, in the kind of "daisy-wheel" typewriters (or printers) the wheel spins around and is eventually struck.  Of course, in both of these cases "you" don't necessarily know from where the ball or wheel came.  
And in the case of modern, computer keyboards, what noises is the system looking at?  The noises the keys themselves might be fairly similar; maybe it is listening for the difference in timing that a human-generated keystroke would exhibit.
At first thought, I figure that using Tempest-type (RF emission) techniques on keyboards would be more plausible:  The encoder chip which scans keys will likely have precise timings.  And it should be possible to program a different encoder chip with intentional emissions of identifiable signals, which can be listened to from a relatively long distance away, tens or hundreds of feet.
                 Jim Bell

  
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/html
Size: 3846 bytes
Desc: not available
URL: <https://lists.cpunks.org/pipermail/cypherpunks/attachments/20190825/d9836848/attachment.txt>


More information about the cypherpunks mailing list