Skip to content
This repository has been archived by the owner on Mar 11, 2024. It is now read-only.

Breaking unlinkability in CL Signatures by colluding Issuers and Verifiers using malicious keys

Low
swcurran published GHSA-2q6j-gqc4-4gw3 Jan 16, 2024

Package

cargo libursa (Rust)

Affected versions

< 0.3

Patched versions

None

Description

CL Signatures Issuer Key Correctness Proof lacks of prime strength checking

A weakness in the Hyperledger AnonCreds specification that is not mitigated in the Ursa and AnonCreds implementations is that the Issuer does not publish a key correctness proof demonstrating that a generated private key is sufficient to meet the unlinkability guarantees of AnonCreds. A sufficient private key is one in which it's components p and q are safe primes, such that:

  • p and q are both prime numbers
  • p and q are not equal
  • p and q have the same, sufficiently large, size
    • For example, using two values both 1024 bits long is sufficient, whereas using one value 2040 bits long and the other 8 bits long is not.

The Ursa and AnonCreds CL-Signatures implementations always generate a sufficient private key. A malicious issuer could in theory create a custom CL Signature implementation (derived from the Ursa or AnonCreds CL-Signatures implementations) that uses weakened private keys such that presentations from holders could be shared by verifiers to the issuer who could determine the holder to which the credential was issued.

Impact

This vulnerability could impact holders of AnonCreds credentials implemented using the CL-signature scheme in the Ursa and AnonCreds implementations of CL Signatures.

Mitigations

Jan Camenisch and Markus Michels. Proving in zero-knowledge that a number is the product of two safe primes (pages 12-13) demonstrates a key correctness proof that could be used to show the issuer has generated a sufficiently strong private key, proving the characteristics listed above.

In a future version of AnonCreds, the additional key correctness proof could be published separately or added to the Credential Definition. In the meantime, Issuers in existing ecosystems can share such a proof with their ecosystem co-participants in an ad hoc manner.

The lack of such a published key correctness proof allows a malicious Issuer to deliberately generate a private key that lacks the requirements listed above, enabling the Issuer to perform a brute force attack on presentations provided to colluding verifiers that breaks the unlinkability guarantee of AnonCreds.

Severity

Low

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v3 base metrics

Attack vector
Local
Attack complexity
Low
Privileges required
None
User interaction
Required
Scope
Unchanged
Confidentiality
Low
Integrity
None
Availability
None

CVSS v3 base metrics

Attack vector: More severe the more the remote (logically and physically) an attacker can be in order to exploit the vulnerability.
Attack complexity: More severe for the least complex attacks.
Privileges required: More severe if no privileges are required.
User interaction: More severe when no user interaction is required.
Scope: More severe when a scope change occurs, e.g. one vulnerable component impacts resources in components beyond its security scope.
Confidentiality: More severe when loss of data confidentiality is highest, measuring the level of data access available to an unauthorized user.
Integrity: More severe when loss of data integrity is the highest, measuring the consequence of data modification possible by an unauthorized user.
Availability: More severe when the loss of impacted component availability is highest.
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:L/I:N/A:N

CVE ID

CVE-2022-31021

Weaknesses

No CWEs