No, the attack would run as follows: 1. I send you a message which is hashed 2. You send back a message which includes the signature in your hash 3. I take the other message with the bad signature to court and your signature on the hash confirming you received it. Jim From: John Bradley [mailto:ve7jtb@xxxxxxxxxx] Hi Jim, I think I am missing something. If the sender is the attacker and has the mac key, then they can integrity protect any plaintext they like with any hash that is supported.. This seems more like the sender wants to use a different hash than the receiver prefers situation. That seems to me like a different security concern than an attacker taking a existing JWS and using a downgrade on the alg parameter to substitute a less collision resistant hash, in order to substitute a plain text (envelope and body in the Compact case) that has the same hash value as that encrypted by the RSA/EC key. Thanks John B. On Sep 19, 2014, at 11:34 PM, Jim Schaad <ietf@xxxxxxxxxxxxxxxxx> wrote: From: jose [mailto:jose-bounces@xxxxxxxx] On Behalf Of John Bradley For HMAC changing the hash from SHA3 256 to SHA2 256 doesn't really get the attacker much as they still don't know the key to be able to create a appropriate fake plaintext. So simply knowing a value that creates a collision is not sufficient for an attack. [JLS] remember that the attacker can be the sender of the message, in this case the key is known to the attacker. I interpreted Mike's comment on PKCS#1 as being that the signature needs to be verified by the algorithm specified in the "alg" parameter. That is not to say that if in the case of PKCS#1 padding if the OID is not consistent with the value of "alg" that wouldn't be an error. I think that would be a invalid JWS. I do think that should be made clear. With PSS if the key is the same length it is true that you are going to be vulnerable to collisions over the plaintext based on the weakest hash you can get the receiver to accept. This is not a issue unique to JOSE in any way. One would be tempted to rule out SHA1 now but it is differentiated by it's key length, so can't be downgraded to from SHA2. John B. On Sep 19, 2014, at 7:25 PM, Jim Schaad <ietf@xxxxxxxxxxxxxxxxx> wrote: From: Richard Barnes [mailto:rlb@xxxxxx] """ In some usages of JWS, there is a risk of algorithm substitution attacks, in which an attacker can use an existing signature value with a different signature algorithm to make it appear that a signer has signed something that he actually has not. These attacks have been discussed in detail in the context of CMS {{RFC 6211}}. The risk arises when all of the following are true:
* Given an existing signature, an attacker can find another payload that produces the same signature value with a weaker algorithm * In particular, the payload crafted by the attacker is valid in a given application-layer context For example, suppose a verifier is willing to accept both "PS1" and "PS256" as "alg" values, and a signer creates a signature using "PS256". If the attacker can craft a payload that has the same SHA-1 digest has as the SHA-256 digest of the legitimate payload, then the "PS1" signature over the bogus payload will be the same as the "PS256" signature over the legitimate payload. There are several ways for an application using JOSE to mitigate algorithm substitution attacks: * Don't accept signatures using vulnerable algorithms: Algorithm substitution attacks do not arise for all signature algorithms. [JLS] This is not a true statement. If you support both SHA256 and SHA512/256 then the signature values have the same length. This will also be an issue if you support both the SHA2 and the SHA3 algorithm sets as they have results of the same length. * The only algorithms defined in JWA {{I-D.ietf-jose-json-web-algorithms}} that is vulnerable to algorithm substitution attacks is RSA-PSS ("PS1", "PS256", etc.). An implementation that does not support RSA-PSS is not vulnerable to algorithm substitution attacks. [JLS] ECDSA is open to this attack if you support both SHA256 and SHA512/256 the hash lengths are the same. This will also be an issue if you support both the SHA2 and the SHA3 algorithm sets as they have results of the same length. * Require that the "alg" parameter be carried in the protected header. (This is the approach taken by RFC 6211.) * Include a field reflecting the algorithm in the application payload, and require that it be matched with the "alg" parameter during verification (This is the approach taken by PKIX {{RFC5280}}.) [JLS] RSA-PKCS#1.5 is open to this attack if the suggestion of Mike in a message prior to this is put into the text. That is to allow for the hash inside of the signature and that outside of the signature to differ in value because you only enforce one of the two values. Of these mitigations, the only sure solution is the first. Signing over the "alg" parameter (directly or indirectly) only makes the attacker's work more difficult, by requiring that the bogus payload also contain bogus information about the signing algorithm. They do not prevent attack by a sufficiently powerful attacker. """ On Fri, Sep 19, 2014 at 2:49 PM, Mike Jones <Michael.Jones@xxxxxxxxxxxxx> wrote: I would appreciate it if you would write a draft of the proposed security considerations text, Richard. Perhaps title the section “Unsecured Algorithm Values”. Thanks! -- Mike From: Richard Barnes [mailto:rlb@xxxxxx]
Richard Barnes writes: I'm perfectly happy to have it documented in the Security Considerations. Mike: Should I generate some text, or do you want to take a stab? _______________________________________________ |