[android-security-discuss] How does "Requiring user authentication in key use" work with compromised Android image?

[Ashish Bhatia]

Ashish
Consider this scenario

1. Android platform on the phone is trustworthy
2. Secure Element is available to store the key in the hardware
3. App A puts the key in the hardware with the fingerprint
authentication requirement
<https://developer.android.com/training/articles/keystore#UserAuthentication>
4. Android platform gets compromised
5. An attacker cannot extract the key from the Secure Element

In this case, can an attacker make the key in Secure Element sign anything
without user interaction? Or, in other words, where is the fingerprint
authentication constraint being verified? Does that happen in the
compromised Android platform image?

Regards,
Ashish

['Brian Carlstrom' via Android Security Discussions]

You might want to look at the "SafetyNet Attestation API" for signals
on compromised devices:

https://developer.android.com/training/safetynet/attestation

obviously nothing is perfect once the platform is compromised.

-bri

Post by Ashish Bhatia
Ashish
Consider this scenario
1. Android platform on the phone is trustworthy
2. Secure Element is available to store the key in the hardware
3. App A puts the key in the hardware with the fingerprint
authentication requirement
<https://developer.android.com/training/articles/keystore#UserAuthentication>
4. Android platform gets compromised
5. An attacker cannot extract the key from the Secure Element
In this case, can an attacker make the key in Secure Element sign anything
without user interaction? Or, in other words, where is the fingerprint
authentication constraint being verified? Does that happen in the
compromised Android platform image?
Regards,
Ashish
--
You received this message because you are subscribed to the Google Groups
"Android Security Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at
https://groups.google.com/group/android-security-discuss.
For more options, visit https://groups.google.com/d/optout.

['Shawn Willden' via Android Security Discussions]

Post by 'Brian Carlstrom' via Android Security Discussions
You might want to look at the "SafetyNet Attestation API" for signals
https://developer.android.com/training/safetynet/attestation
obviously nothing is perfect once the platform is compromised.

Just to clarify, compromising the platform does not compromise the
integrity of fingerprint authentication or the binding of keys to
authentication. All of that is done in the Trusted Execution Environment
(TEE). Given that the TEE has a much, much smaller attack surface than the
platform, it should be correspondingly harder for an attacker to break. TEE
implementations should also receive more scrutiny than the platform since
they're both more critical to device security and smaller. On the other
hand, they're also mostly closed source which tends to deter (but not
prevent) vulnerability research.

Post by 'Brian Carlstrom' via Android Security Discussions
-bri

Post by Ashish Bhatia
Ashish
Consider this scenario
1. Android platform on the phone is trustworthy
2. Secure Element is available to store the key in the hardware
3. App A puts the key in the hardware with the fingerprint
authentication requirement
<https://developer.android.com/training/articles/keystore#UserAuthentication>
4. Android platform gets compromised
5. An attacker cannot extract the key from the Secure Element
In this case, can an attacker make the key in Secure Element sign
anything without user interaction? Or, in other words, where is the
fingerprint authentication constraint being verified? Does that happen in
the compromised Android platform image?
Regards,
Ashish
--
You received this message because you are subscribed to the Google Groups
"Android Security Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an
<javascript:>.
Visit this group at
https://groups.google.com/group/android-security-discuss.
For more options, visit https://groups.google.com/d/optout.

[Ashish Bhatia]

Thanks a lot, Shawn and Brian for the prompt response.


Ashish

On Wed, Oct 3, 2018 at 1:54 PM 'Shawn Willden' via Android Security

Post by 'Shawn Willden' via Android Security Discussions

Post by 'Brian Carlstrom' via Android Security Discussions
You might want to look at the "SafetyNet Attestation API" for signals
https://developer.android.com/training/safetynet/attestation
obviously nothing is perfect once the platform is compromised.

Just to clarify, compromising the platform does not compromise the
integrity of fingerprint authentication or the binding of keys to
authentication. All of that is done in the Trusted Execution Environment
(TEE). Given that the TEE has a much, much smaller attack surface than the
platform, it should be correspondingly harder for an attacker to break. TEE
implementations should also receive more scrutiny than the platform since
they're both more critical to device security and smaller. On the other
hand, they're also mostly closed source which tends to deter (but not
prevent) vulnerability research.

Post by 'Brian Carlstrom' via Android Security Discussions
-bri

Post by Ashish Bhatia
Ashish
Consider this scenario
1. Android platform on the phone is trustworthy
2. Secure Element is available to store the key in the hardware
3. App A puts the key in the hardware with the fingerprint
authentication requirement
<https://developer.android.com/training/articles/keystore#UserAuthentication>
4. Android platform gets compromised
5. An attacker cannot extract the key from the Secure Element
In this case, can an attacker make the key in Secure Element sign
anything without user interaction? Or, in other words, where is the
fingerprint authentication constraint being verified? Does that happen in
the compromised Android platform image?
Regards,
Ashish
--
You received this message because you are subscribed to the Google
Groups "Android Security Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send
Visit this group at
https://groups.google.com/group/android-security-discuss.
For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to the Google Groups
"Android Security Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at
https://groups.google.com/group/android-security-discuss.
For more options, visit https://groups.google.com/d/optout.

['Shawn Willden' via Android Security Discussions]

Fingerprint authentication is done in the Trusted Execution Environment
(TrustZone in ARM devices).

See https://source.android.com/security/authentication

Post by Ashish Bhatia
Ashish
Consider this scenario
1. Android platform on the phone is trustworthy
2. Secure Element is available to store the key in the hardware
3. App A puts the key in the hardware with the fingerprint
authentication requirement
<https://developer.android.com/training/articles/keystore#UserAuthentication>
4. Android platform gets compromised
5. An attacker cannot extract the key from the Secure Element
In this case, can an attacker make the key in Secure Element sign anything
without user interaction? Or, in other words, where is the fingerprint
authentication constraint being verified? Does that happen in the
compromised Android platform image?
Regards,
Ashish
--
You received this message because you are subscribed to the Google Groups
"Android Security Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at
https://groups.google.com/group/android-security-discuss.
For more options, visit https://groups.google.com/d/optout.