[Cryptography] The FBI can (almost certainly) crack the San Bernardino iPhone without Apple's help

Natanael natanael.l at gmail.com
Thu Mar 3 08:20:50 EST 2016


Den 3 mar 2016 14:12 skrev "Perry E. Metzger" <perry at piermont.com>:
>
> On Thu, 3 Mar 2016 01:25:26 +0100 Natanael <natanael.l at gmail.com>
> wrote:
> > The solution is to let us choose our walled garden, including the
> > option of building our own. Let us be our own root authority if we
> > wish. Most might prefer the factory default, but the option to
> > switch mode to insert our own root so that we can control every
> > security setting and every behavior should be there for us.
>
> If the option is available and simple enough to turn on, then
> phishermen will find ways to induce naive users to turn it on, and
> people doing supply chain attacks or "evil maid" attacks will turn it
> on when they intercept equipment.
>
> On the other hand, although I suspect that (say) Apple will do a
> better job securing their equipment than I will, there are a
> number of manufacturers of more specialized equipment for whom this is
> *not* true and I want the ability, as a sophisticated user, to
> tighten the security on their systems.
>
> So, the situation is not cut and dried, and there are few obvious
> clean answers. We live in a dangerous world.

On Android, bootloader unlocking wipes the phone. Having a static lowest
level bootloader and TPM that keeps track of the active mode and chosen
root can enforce this. Something as simple as a hardware token for
enrolling could make it easy - when you get the phone, boot to
reconfiguration mode and plug in your hardware token, load the root key.
The phone resets and now uses your key.

Want to use the factory key again? Follow a simple guide to do a factory
reset from the bootloader, and it resets to its original state directly.
Telling users to always do this with new phones is really the best we can
do.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20160303/155c821c/attachment.html>


More information about the cryptography mailing list