As debate intensifies over whether artificial intelligence will accelerate breakthroughs that threaten the cryptographic systems used by cryptocurrencies and the broader internet, related development efforts are advancing.
Falcon (Fast-Fourier Lattice-based Compact Signatures over NTRU) is a lattice-based digital signature scheme designed to resist attacks from quantum computers. Researchers submitted Falcon to NIST’s Post-Quantum Cryptography project in 2017.
The team has integrated Falcon-512 into its key management interface, while other post-quantum options, including ML-DSA-44, Winternitz, and XMSS, remain marked as "coming soon."
Mixed signatures may enhance multisignature security
Squads states that its solution allows multisig participants to use different signing schemes within the same approval structure. For example, in a 2-of-3 multisig, if two signers use Falcon-512 and one uses the traditional Ed25519, an attacker who compromises the Ed25519 signer would still need to compromise at least one Falcon-512 signer to meet the approval threshold. This configuration introduces cryptographic diversity without requiring all participants to migrate to the new signing scheme.
However, this implementation does not make every part of Solana transactions post-quantum resistant. Even though Falcon-512 handles multi-signature authorization, Solana still requires the transaction fee payer to provide a native Ed25519 signature.
Anatoly Yakovenko urges developers to prepare in advance.
This development follows a call on October 8 by Solana co-founder Anatoly Yakovenko, who urged wallet developers to begin implementing post-quantum cryptography and establish robust processes for managing and rotating cryptographic keys.
Previously, Ethereum Foundation researcher Justin Drake warned that the crypto industry should consider implementing defensive measures to address potential breakthroughs from AI-driven mathematics.
Drake believes that attackers may crack existing cryptography earlier than expected. Other researchers have questioned this urgency, noting that advances in AI mathematics do not imply that attackers can solve the difficult mathematical problems underlying existing signature schemes.
There is currently no public evidence that AI has broken the core cryptographic assumptions underlying widely used digital signatures.
Just an early step, not a complete solution.
The broader challenge is not just selecting a post-quantum algorithm; developers must also establish secure key management, reliable migration processes, and robust implementations capable of withstanding real-world conditions.