A collaborative effort among over 100 researchers utilizing AI coding agents has resulted in an 86% reduction in the cost of quantum attacks targeting Bitcoin and Ethereum. This significant advancement, which lowered the benchmark for breaking Bitcoin's cryptography from 10.75 billion to 1.496 billion, was achieved between late May and July 26, with results published recently.
Bitcoin and Ethereum rely on the secp256k1 elliptic curve to secure signatures, making them vulnerable to potential quantum computing threats. The ECDSA.Fail competition, organized by Eigen Labs, assessed circuit designs based on the number of logical qubits and Toffoli gates, an expensive quantum operation. The most efficient design used 1,151 logical qubits and about 1.3 million Toffoli gates, while subsequent designs achieved even lower gate counts, significantly reducing resource needs.
This research, supported by organizations such as Eigen Labs, Trail of Bits, StarkWare, and the Ethereum Foundation, emphasizes the importance of understanding the costs associated with quantum threats to better plan defenses. Their approach, termed Open Autoresearch, combines human expertise with AI to achieve measurable goals in quantum security.
