[Rats] Re: Follow-up of interim: Jun's and Thomas' comments about two keys

[Muhammad Usama Sardar <[e-mail]>]

Hi Thomas,

I believe the email is pretty clear. For details, I have referred to the 
paper. If you want to go to details, please be precise which exact 
statement in the paper is your concern related to? Presenting high-level 
concerns without pointing to the exact statement in the paper is not 
useful at all.

On 11.02.26 10:58, Thomas Fossati wrote:
> I still fail to understand the scope of this conversation.
Paper is clearly outlining the scope, and has been discussed several 
times in CCC Attestation SIG as well as RATS.What exactly confuses you?
> Once the CC platform is broken -- as is the fundamental assumption
> here (see the mention to TEE.fail & co.) -- and the verifier is
> informed of this, the platform measurements will contain evidence that
> the peer is unreliable. So I believe this is not a scenario you want
> to protect against, is it?
How can the verifier know who owns the exact instance of the 
platform?Verifier also needs Endorsements of which platforms belong to 
the data centers of the desired CSP. To the best of our knowledge, no 
cloud provider currently publishes such Endorsements.
> It seems to me that the scope is a zero-day and its ramifications in
> the context of TLS connections being relayed by an active attacker to
> a broken CC node, where the verifier is not yet aware of the broken
> TCB.  Is this correct?
No, that's not correct. The existing protocols and implementations 
provide no way to identify an instance belongs to a specific cloud 
provider or someone's basement. In short, a machine in someone's 
basement anywhere in the world is indistinguishable from a specific 
cloud provider's machine in a desired region.
> In general, I think you need to make it much clearer to your audience
> what the applicability of this attack is, because at the moment, one
> is left a bit wondering.

Audience is presented with a paper as well as the repo, which has links 
to the several talks where all of these issues have already been 
addressed to the satisfaction of everyone. If you have any leftover 
issue, please be precise and point to the exact statement in the paper.

Thanks,

-Usama