ML-DSA
A post-quantum digital signature algorithm standardized by NIST, based on the module learning with errors problem. It is used to replace traditional signature schemes such as RSA and ECDSA.
Core Facts
Timeline (last 90 days)
Post-quantum AI infrastructure should adopt ML-KEM (Kyber) to protect inter-cluster communication and model distribution channels, replace ECDSA signatures with ML-DSA (Dilithium), and build a crypto-agile agent framework.
FIPS 204 (ML-DSA) is derived from CRYSTALS-Dilithium and is used for digital signatures.
The security of ML-KEM and ML-DSA is built on mathematical hard problems such as the 'shortest vector problem on lattices,' which to date are both quantum-resistant and have not been significantly accelerated by AI-based attacks.
ML-KEM and ML-DSA use Module-LWE as their core construction; Module-LWE is a structured variant of LWE that improves computational efficiency and reduces key sizes.
ML-KEM was formerly known as CRYSTALS-Kyber and is a lattice-based key encapsulation mechanism; ML-DSA was formerly known as CRYSTALS-Dilithium and is a lattice-based digital signature algorithm; SLH-DSA was formerly known as SPHINCS+ and is a hash-based digital signature scheme.
The OpenJDK community has begun planning JEPs aimed at integrating NIST standard algorithms such as ML-KEM and ML-DSA as native implementations into the standard library.
All Facts (7)
NIST formally published three post-quantum cryptography Federal Information Processing Standards (FIPS) in August 2024.
65%UnverifiedPost-quantum AI infrastructure should adopt ML-KEM (Kyber) to protect inter-cluster communication and model distribution channels, replace ECDSA signatures with ML-DSA (Dilithium), and build a crypto-agile agent framework.
50%UnverifiedFIPS 204 (ML-DSA) is derived from CRYSTALS-Dilithium and is used for digital signatures.
50%UnverifiedThe security of ML-KEM and ML-DSA is built on mathematical hard problems such as the 'shortest vector problem on lattices,' which to date are both quantum-resistant and have not been significantly accelerated by AI-based attacks.
50%UnverifiedThe OpenJDK community has begun planning JEPs aimed at integrating NIST standard algorithms such as ML-KEM and ML-DSA as native implementations into the standard library.
50%UnverifiedML-KEM and ML-DSA use Module-LWE as their core construction; Module-LWE is a structured variant of LWE that improves computational efficiency and reduces key sizes.
50%UnverifiedML-KEM was formerly known as CRYSTALS-Kyber and is a lattice-based key encapsulation mechanism; ML-DSA was formerly known as CRYSTALS-Dilithium and is a lattice-based digital signature algorithm; SLH-DSA was formerly known as SPHINCS+ and is a hash-based digital signature scheme.
50%