AI-Fueled Breakthroughs in Cryptography Raise Concerns for Bitcoin Security

The research firm Anthropic has made significant discoveries in the field of cryptography using its AI tool, Claude Mythos Preview model. The model identified vulnerabilities in cryptographic algorithms that accelerate the breaking of reduced versions of the AES algorithm by 200 to 1,000 times and halved the key strength in HAWK, a candidate signature scheme for NIST's post-quantum standardization.

The researchers used the AI tool to discover new mathematical vulnerabilities in encryption schemes before quantum computers could compromise current signatures on networks like Bitcoin. The model was able to develop an attack method against a simplified seven-round version of the Advanced Encryption Standard (AES-128) and accelerated the breaking process.

The results suggest that frontier models are acquiring capabilities to generate autonomous and unprecedented cryptographic research. The same model structured an improved attack against HAWK, requiring approximately 60 hours of processing and costing nearly $100,000 in API consumption. According to data released by Anthropic, the operation managed to halve the effective key strength in the HAWK-256 test parameter.

Bitcoin's current architecture employs the elliptic curve digital signature algorithm (ECDSA) to authenticate ownership of funds and validate blocks. Research reports suggest that quantum processors with a sufficient number of physical qubits could compromise the security of ECDSA in the long term. The emergence of post-quantum algorithms is being considered as a potential solution.

Peter Van Valkenburgh, Executive Director of Coin Center, stated that these advances present a scenario where developers must consider both the progress of quantum hardware and the advancement of artificial intelligence algorithms when evaluating the security of new standards.