[SSL Observatory] DigiNotar Certificate Hierarchy

Jacob Appelbaum jacob at appelbaum.net
Tue Sep 6 04:12:06 PDT 2011


On 09/06/2011 12:44 PM, Peter Gutmann wrote:
> Gervase Markham <gerv at mozilla.org> writes:
> 
>> I am attempting to chart the DigiNotar certificate hierarchy in a public
>> document, so Mozilla can be sure that our current block is sufficiently wide.
> 
> In case you haven't seen the Comodo, possibly Diginotar, and now StartSSL,
> Globalsign, and several others hacker's post:
> 
>   http://pastebin.com/1AxH30em
> 
> you may need to get ready to block a whole lot more than just Diginotar in the
> near future.  Unlike Diginotar, we're now getting into the TB2F CAs, so the
> response will be interesting...

I think this is a pretty bad place to get and it seems pretty obvious
that we've arrived. I hope that we don't need these guys to screw up a
third time before we learn this lesson in deployed software. The Tor
Project pinned our certs in Chrome after the first time we were directly
targeted, we had some internal debates after ComodoGate and the
DigiNotar Debacle pretty much ended the debate.

I really wish it was possible for users of Firefox to be similarly
secure but it sure doesn't seem possible at this point. Hoping that
policy discussions end well seems like the best strategy and that isn't
much of a security plan...

I sure hope the Browser vendors are really ready to pull *any* and
*every* CA in their trust root if there is evidence of a compromise;
regardless of the hilariously successful "Iranian" hackers, I think
issuing certs to law enforcement for surveillance is essentially the
same level of compromise from the user standpoint. I hope future
policies will reflect this concern... though I'm not going to hold my
breath...

All the best,
Jacob



More information about the Observatory mailing list