1 min read
Add as a preferred source on Google

Dragonfly Partner Says Ethereum ‘Bunker Mode’ Won’t Fully Solve AI Wallet Risk

Haseeb Qureshi said moving assets to new addresses would not prevent attacks on other wallets from causing losses if compromised tokens were sold heavily.

Mentioned assets
Two hardware wallets and a blank recovery card on a table / TokenPost.ai
Two hardware wallets and a blank recovery card on a table / TokenPost.ai

Dragonfly managing partner Haseeb Qureshi said Ethereum researcher Justin Drake’s proposed “bunker mode” would not fully address the risk that artificial intelligence could eventually break cryptographic wallet signatures.

Qureshi said moving assets to new addresses would protect assets that remained there, but would not eliminate the broader market risk. Tokens held in other wallets could still lose value if they were attacked and sold in large amounts.

The debate follows Drake’s warning that rapid advances in AI’s mathematical capabilities could put the Elliptic Curve Digital Signature Algorithm, or ECDSA, at risk within “months rather than years.” ECDSA is used to protect crypto wallets and authorize transactions.

Drake has urged users to gradually move funds to new wallets whose public keys have not been exposed. His broader warning and preparation plan were outlined in earlier coverage of ECDSA preparations.

Qureshi proposed a different approach called “Cryptographic Recovery Mode.” The framework would use hash-based backup signatures, with users binding recovery mechanisms to their wallets in advance. If existing cryptographic signatures were later broken, validators could help restore access to affected assets.

The competing proposals reflect different approaches to a future wallet-security threat: reducing exposure through new addresses or preparing an advance recovery process that could be activated after a signature system fails.

Simon Yoon

Reporter

Simon Yoon reports on blockchain technology for TokenPost. Send corrections or tips to info@tokenpost.com.

Loading…