AI Agents Slash Quantum Attack Costs for Bitcoin by 86%
A team of over 100 researchers utilized AI coding agents to reduce the resource score for a crucial step in a quantum attack against Bitcoin by 86.1%, according to a paper published on September 9.
The optimization, called ECDSA.Fail, targeted secp256k1 point addition circuits and lowered the score from 10.75 billion to 1.496 billion. The researchers emphasized that their work does not represent a breakthrough in hacking Bitcoin or recovering private keys but rather an example of how software optimization can decrease quantum resources needed for elliptic-curve arithmetic.
Experts warn, however, that while this development is significant, it may only provide 'breathing room' as the real challenge lies in governance and migration. The researchers estimate that a full secp256k1 attack would still require a machine that does not yet exist.