AI Reduces Quantum Bitcoin Attack Costs Significantly

A quantum computer capable of cracking Bitcoin’s cryptography has long been treated as a distant, almost theoretical problem. But new research suggests the math behind that threat is shrinking faster than anyone expected, and artificial intelligence is the reason why. Over just two months, a group of researchers working alongside AI coding agents managed to reduce a key benchmark for a quantum attack on Bitcoin and Ethereum by 86%, according to a paper published this week and first reported by Decrypt. The finding shows how AI reduces quantum Bitcoin attack costs in ways that could reshape the urgency around blockchain security.

Key takeaways

• AI coding agents helped researchers cut a quantum attack benchmark on Bitcoin and Ethereum from 10.75 billion to 1.496 billion operations, an 86% drop, in just two months.
• More than 100 researchers from Eigen Labs, Trail of Bits, StarkWare, Theta Labs, MultiVM Labs and the Ethereum Foundation contributed to the work.
• The Ethereum Foundation still targets December 2029 for quantum-resistant transactions, validators and storage, even as attack-side research accelerates.
• Binance’s Bitcoin reserves hit a two-year high near 693,000 BTC, while Bitcoin ETFs logged $283 million in net outflows and Ethereum ETFs saw $30 million leave.