Justin Drake, a researcher at the Ethereum Foundation, warns that artificial intelligence could potentially compromise the ECDSA encryption used by Bitcoin and Ethereum. He suggests that AI might achieve this within months, potentially outpacing the threat posed by quantum computing.
Drake advises the blockchain industry to initiate a controlled migration of assets to fresh addresses that have never performed a transaction. This strategy, termed bunker mode, aims to keep public keys hidden behind hashes, as public keys become visible on the blockchain only after a transaction is sent.
The researcher notes that AI is rapidly solving complex mathematical problems, citing 722 mathematical results recently published by OpenAI. He argues that the elliptic curves underlying ECDSA contain patterns that advanced systems could exploit to recover private keys using large clusters of graphics cards.
Drake has urged major industry players, including Binance, Bitbank, Robinhood, Bitfinex, and Tether, to strengthen their cold storage security. He suggests that attackers would likely target high-profile wallets first, such as those associated with Bitcoin's creator, which hold 50 BTC each.
For individual users, Drake recommends moving self-custodied assets to new, secure wallets or utilizing established custodians like Coinbase and ETFs. He emphasizes that users should avoid panic and instead focus on implementing best practices for asset protection.
