Very nice synopsis Shawn. It seems
clear to me that NIST would prefer a TPM solution. One of the
authors of the draft standard spoke to the attendees of a meeting
at NIST last Friday (Mobile Device Security Technical Exchange)
and emphasized that preference including in phones and tablets.
That is nothing new however.
On 12/18/12 11:59 AM, Shawn Geddis wrote:
On Dec 18, 2012, at 2:25 AM, Jeffrey Walton <email@hidden>
wrote:
I'm question TIA's claims that a
smartphone is like a laptop (top of page two). When's the last
time you heard an encrypted laptop hard drive leaked passwords
and other secrets like an iPhone's KeyChain?
There is definitely a gap in security controls. Confer:
http://sit.sit.fraunhofer.de/studies/en/sc-iphone-passwords-faq.pdf.
Jeffrey,
Maybe you didn't understand that the document covers iPhone
OS 3.x thru iOS 6.0.1 ? Protections have changed in raising the
bar without negatively impacting usability.
Page 9:
Is
the iOS keychain in general insecure?
No,
the iOS keychain can provide sufficient security
for stored items, if both of
the following conditions are met:
•
Items are stored with a protection class that
makes them only available
when the device is unlocked. (see Section 2.14)
• A
strong passcode of 6 alphanumeric digits is
enforced (reduces the risk1
of brute-force attacks).
Page 6:
However,
in iOS 3.x and prior versions all entries of the
keychain can be accessed by attackers. In these
versions all keychain entries are encrypted only
with the device key accessible on the device.
Therefore, for those devices, the
knowledge of the passcode is not needed to decrypt
all keychain entries.
NOTE: Enhanced Data Protection Framework was not
introduced until iOS 4.0. The DEFAULT protection for
files within Apps created prior to iOS 4.0 and those not
specifying enhanced data protection is referred to as
"ProtectionNone" ("Class D") -- Class Keys can be unlocked and
accessed solely with the OS leveraging the Device Key upon
boot without human interaction.
In other words, if you use what Apple provides in the OS,
they confirm the sufficient security. If you choose not to use
the built-in services to protect your data, it equates to
shooting yourself in the foot and then wondering why.
You can lead a horse to
water, but....
As for the SP-800-164 DRAFT posted by NIST.....
The article claims:
"suggests that security in mobile devices can only
be realized using a specific architectural implementation of
secure or trustworthy environment, namely the Trusted
Platform Module (TPM) architecture specified by the Trusted
Computing Group (TCG).
When in fact it does not say that at all. What it says is
(pg 5 is first example):
Current mobile devices lack the hardware-based root
of trust features that are increasingly built into laptops and
other types of hosts (e.g., Trusted Platform
Modules, TPMs).
TPM was noted as an example, but that is clearly not
stated as the only approach to providing a HW RoT.
iOS devices in-fact do contain a Hardware Root of Trust:
Note that Apple does not use TPMs. Etched in hardware,
immutable, are three objects that constitute a Hardware Root
of Trust (RoT):
* Apple Root CA Certificate (used for signature
validation in Secure Boot Chain)
* Group Key (unique to a Product Family - e.g. 'iPhone
5')
* Device Key (unique to the device and unknown to
the OS / Apps - think black box)
- Shawn
________________________________________
Shawn Geddis
Security Consulting Engineer
Apple Enterprise Division