Post-Quantum Vulnerable
Documentation: Post-Quantum Vulnerable SSL and Recommended Countermeasures
Overview
The rise of quantum computing poses a direct threat to traditional cryptographic algorithms used in SSL/TLS, such as RSA, ECDSA, and Elliptic Curve Diffie-Hellman (ECDH). These algorithms rely on problems (like integer factorization and discrete logarithms) that quantum computers can solve efficiently using Shorβs algorithm, rendering current SSL/TLS protocols post-quantum vulnerable.
This document outlines how SSL/TLS can be vulnerable in a post-quantum context and provides recommendations for post-quantum secure alternatives.
β οΈ Post-Quantum Vulnerabilities in SSL/TLS
1. Key Exchange Vulnerabilities
Traditional key exchange mechanisms such as:
-
RSA key exchange
-
Elliptic Curve Diffie-Hellman (ECDH)
These are vulnerable because quantum computers can derive the private keys from intercepted public data using Shorβs algorithm.
2. Digital Signature Vulnerabilities
-
RSA and ECDSA signatures can be forged with quantum algorithms.
-
Intercepted SSL traffic today can be decrypted in the future using harvest-now, decrypt-later attacks.
π Modern Post-Quantum Safe Ciphers (Hybrid and Pure)
To mitigate quantum risks, TLS implementations must transition to quantum-resistant cryptographic algorithms. These can be:
-
Hybrid schemes: Combine classical and post-quantum algorithms (interoperable and backward compatible)
-
Pure PQC schemes: Use only post-quantum algorithms (ideal for future-proofing)
π NIST-Recommended Post-Quantum Algorithms (2024)
Type | Algorithm | Description |
---|---|---|
Key Encapsulation | Kyber (Kyber-512, Kyber-768, Kyber-1024) | Lattice-based; standardized as ML-KEM |
Signatures | Dilithium (Dilithium2, Dilithium3, Dilithium5) | Lattice-based digital signature |
Β | Falcon | Compact, lattice-based signature |
Β | SPHINCS+ | Hash-based signature; conservative fallback |
β Recommended Cipher Suites (TLS 1.3 with PQC)
As of today, post-quantum cryptography is often integrated through TLS hybrid key exchanges via extensions like [x25519+kyber512] or via Open Quantum Safe (OQS) libraries.
Example Hybrid Cipher Suites (using NIST PQC algorithms):
-
TLS_AES_256_GCM_SHA384 + X25519+Kyber768
-
TLS_CHACHA20_POLY1305_SHA256 + Kyber512
-
TLS_AES_128_GCM_SHA256 + ECDHE + Kyber1024
β οΈ These are not yet formal cipher suite names; they represent the key exchange combinations used in hybrid TLS libraries like BoringSSL, OpenSSL with OQS, and AWS s2n-tls.
π§° Implementation Options
Libraries supporting Post-Quantum TLS:
-
-
Extensions for OpenSSL, BoringSSL, and liboqs
-
-
Google BoringSSL
-
Early experimental support for hybrid schemes
-
-
AWS s2n-tls
-
Supports hybrid post-quantum TLS 1.2 and TLS 1.3
-
-
Cloudflare and Google Chrome (experimental support in QUIC)
π Migration Strategy
-
Evaluate inventory of TLS endpoints
-
Identify services using RSA/ECDSA and vulnerable ciphers.
-
-
Use hybrid schemes first
-
Interoperable with legacy clients while offering PQC protection.
-
-
Enable PQC cipher suites in controlled environments
-
Internal APIs, backend services, etc.
-
-
Monitor NIST and IETF standards
-
Stay aligned with official PQC cipher standardization efforts.
-
-
Implement TLS 1.3
-
Required for most PQC-ready key exchange implementations.
-
π« Ciphers to Deprecate Immediately
Avoid these in future-proof systems:
-
TLS_RSA_WITH_AES_256_GCM_SHA384
-
TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA
-
TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384
-
TLS_DHE_RSA_WITH_AES_256_CBC_SHA
These are based on RSA/ECDSA and vulnerable to quantum attacks.
NIST PQC Standardization Milestones
Year | Milestone |
---|---|
2022 | NIST announced finalists (Kyber, Dilithium, etc.) |
2024 | Final standards (ML-KEM for Kyber, ML-DSA for Dilithium) |
2025 | Adoption into commercial systems and TLS libraries |
Summary
-
SSL/TLS systems that rely on RSA/ECDSA are vulnerable to quantum attacks.
-
Transitioning to post-quantum key exchange and signature schemes is critical.
-
Adopt hybrid schemes like X25519+Kyber during the transition phase.
-
Monitor evolving standards from NIST, IETF, and TLS WG.
No Comments