1

So a few weeks ago I came across a security report that stated that the SHA1 thumbprint of the certificate was a vulnerability. The vulnerability was raised due to fact that the thumbprint used the SHA1 algorithm and this algorithm has structural flaws.

From what I understood, the thumbprint and signature are entirely unrelated, as the thumbprint is not actually a part of the certificate itself, thus there are no security issues. Further research into this topic, confirms my understanding and I believe that there is a misunderstanding by the writer of this report and the vulnerability itself is incorrect.

As a final measure I thought I would ask here to ensure I am not missing anything.

Is it a security concern if a certificate's thumbprint is calculated using the SHA1 algorithm?

1
  • 4
    "... that the SHA1 thumbprint of the certificate was a vulnerability" - The question is missing essential information. In which context was the thumbprint used? If it was just displayed - no problem. If it was used to verify the certificate for example for certificate pinning - problem. Mar 3, 2021 at 15:39

4 Answers 4

3

Is it a security concern if a certificate's thumbprint is calculated using the SHA1 algorithm?

This depends on how the thumbprint is used later.

If the thumbprint is used in an application as the only way to verify a specific certificate then the signature on the certificate does not matter at all, because it is not verified. Instead in this case the security of the certificate is reduced to the security of the SHA-1 thumbprint and therefore weak.

1
  • Thanks for your response Steffen. No the thumbprint is not used later, they are merely raising the fact that the thumbprint is calculated using SHA-1.
    – SecGuy45
    Mar 11, 2021 at 14:14
2

You didn't miss anything. Thumbprint is not a part of the certificate and is calculated on a fly. And having an SHA1 thumbprint isn't a vulnerability.

1
  • Thanks for the response Crypt32. This is my understanding.
    – SecGuy45
    Mar 11, 2021 at 14:15
2

If thumbprint is calculated after the certificate is already signed it should not be a problem. And last I checked, SHA-1 is still resistant to pre-image and second pre-image attacks

0
0

SHA-1 is broken

Let's start with wikipedia/sha-1:

Since 2005, SHA-1 has not been considered secure against well-funded opponents;[4] as of 2010 many organizations have recommended its replacement.[5][6][7] NIST formally deprecated use of SHA-1 in 2011 and disallowed its use for digital signatures in 2013. As of 2020, chosen-prefix attacks against SHA-1 are practical.[8][9] As such, it is recommended to remove SHA-1 from products as soon as possible and instead use SHA-2 or SHA-3. Replacing SHA-1 is urgent where it is used for digital signatures.

You are correct that hashing the message to be signed (in this case a certificate) and actually computing the signature are separate operations, however the signature relies on the hash function being secure; if the hash function is broken, then the whole signature is broken.

The attacks starting in 2005 were collision attacks, meaning that the attacker could do offline computation to compute two certificates that have the same SHA-1 hash -- say one was for secguy45.com and the other for google.com and both with public keys that you have the private key for. You submit the CSR for secguy45.com to a public CA, they happily sign it, but since both certs have the same SHA-1 hash, the signature on that cert also happens to be valid for your google.com cert. Congrats, you are now google.com. (This is a bit simplified since you don't directly submit a cert to a CA, but a CSR, some complexity of the attack omitted...). That was already bad; and the further attack techniques published in 2015, 2017, 2019 just further sink the security of SHA-1.


SHA-1 in certificates

HTTPS certificates for web browsers are governed by a standards body called the CA/Browser Forum, and their core document is called the Baseline Requirements. In 2014 they began sunsetting SHA-1 in certificates, with SHA-1 certificates to be completely gone by Jan 2017.

It's now 2021. Where on earth did you find a still-valid SHA-1 certificate? It's time you got rid of it!

3
  • Mike, you missed entire point. The question was about a Thumbprint, not signature.
    – Crypt32
    Mar 3, 2021 at 15:18
  • @Crypt32 Fair. I'll leave my answer up for now because the question does not make a lot of sense to me; if the security scanner is using SHA-1 to compute the fingerprint and then complaining about its own choice to use SHA-1 ... that's just weird. I agree with Steffen that more information is needed from the OP, it's possibly that the scanner means "thumbprint" as in "the hashing step of signature generation". Mar 3, 2021 at 15:51
  • In my practice I faced flawed security scanners that report false positive alerts and this may be the case.
    – Crypt32
    Mar 3, 2021 at 16:26

Not the answer you're looking for? Browse other questions tagged .