Verisign/RapidSSL Responds To Certificate Vulnerability(blogs.verisign.com)
blogs.verisign.com
Verisign/RapidSSL Responds To Certificate Vulnerability
https://blogs.verisign.com/ssl-blog/2008/12/on_md5_vulnerabilities_and_mit.php
3 comments
"Q: How many certificates are affected?
A: Zero. No end entity certificates are affected by this attack. The attack, when it worked, was a potential method for a criminal to create a new, false certificate from scratch. Existing certificates are not targets for this attack."
This answer seems a little bit, well, disingenuous. Does he not understand the attack or does he not understand SSL?
A: Zero. No end entity certificates are affected by this attack. The attack, when it worked, was a potential method for a criminal to create a new, false certificate from scratch. Existing certificates are not targets for this attack."
This answer seems a little bit, well, disingenuous. Does he not understand the attack or does he not understand SSL?
I believe the answer for that question was actually pulled from the MD5 paper itself. See here:
http://www.win.tue.nl/hashclash/rogue-ca/
Scroll to the bottom in the question/answer section. It's interesting they only used 200 PS3 game consoles to do their computations with.
http://www.win.tue.nl/hashclash/rogue-ca/
Scroll to the bottom in the question/answer section. It's interesting they only used 200 PS3 game consoles to do their computations with.
In the paper itself, they reach the opposite conclusion:
"Question. Suppose that a criminal creates by our method his own rogue Certification Authority that is trusted by all browsers. Are then only websites with certificates from the CA whose signature he used vulnerable to impersonation attacks? Are only websites with certificates based on MD5 vulnerable to impersonation attacks?
Answer. No. When a criminal uses redirection attacks in combination with a rogue but trusted CA certificate, all websites are equally vulnerable."
"Question. Suppose that a criminal creates by our method his own rogue Certification Authority that is trusted by all browsers. Are then only websites with certificates from the CA whose signature he used vulnerable to impersonation attacks? Are only websites with certificates based on MD5 vulnerable to impersonation attacks?
Answer. No. When a criminal uses redirection attacks in combination with a rogue but trusted CA certificate, all websites are equally vulnerable."
(Thanks to gojomo for finding this).
Gotta admit, a solid response.
Gotta admit, a solid response.
Yes, they're taking a page out of Microsoft's playbook by using a blog post to respond in more detail to a security vulnerability. When executed well, it mollifies critics much better than the press release style of communication due to its appearance of informality.
That being said, I would have liked for them to say that they've reviewed their logs and saw no other issuance activity that followed the same pattern as the researchers, and that they will improve their operational monitoring and serial number sequencing to help protect against future potential attacks.
That being said, I would have liked for them to say that they've reviewed their logs and saw no other issuance activity that followed the same pattern as the researchers, and that they will improve their operational monitoring and serial number sequencing to help protect against future potential attacks.
They basically sucked all the drama right out of it, which seems like exactly the right play.
Gotta admit that's pretty much what "a product marketing executive for VeriSign's SSL business unit" should say in a damage control mode.
I really like this, especially the last sentence. That's refreshingly direct communication.