AI-assisted mathematical advances have prompted a debate among Ethereum researchers about the future security of crypto wallets. The concern is serious, but conditional: two leading researchers warn that, in the worst case, the signature system wallets use could be broken within months. The source also makes clear that no one has yet broken the current ECDSA signature scheme in practice.
Why wallet signatures matter
A wallet signature is connected to the keys that control its funds. According to Justin Drake, signing a transaction reveals the public key. An attacker could theoretically use that exposed key to derive the private key and take the funds.
That risk depends on what information is visible. Drake recommends moving funds to addresses that have never signed a transaction. His reasoning is that, when only the public key's hash is visible, the funds remain protected because the key cannot be recovered from that hash.
This makes the distinction between a public key and its hash central to the proposed precaution. The recommendation is not simply to move funds anywhere: it is to use addresses that have not exposed a public key through a signature.
Drake urges preparation
Justin Drake called on the blockchain industry to prepare a “bunker mode” in a post on X. The phrase describes a readiness posture for a possible threat, rather than evidence that wallet signatures have already been compromised.
His warning is framed around a worst-case timeline. AI-assisted math could, in that scenario, make it possible to break the signature system within months. That prospect gives wallet security teams a reason to consider how they would respond, even though the source reports no practical break of ECDSA so far.
The difference between preparing and declaring an immediate emergency matters. Drake's recommendation focuses on reducing exposure by moving funds to addresses that have never signed. The underlying concern is that a public key exposed by signing might eventually become useful to an attacker if mathematical capabilities advance.
Buterin urges care with migrations
Ethereum co-founder Vitalik Buterin agreed in a reply on X, but cautioned against moving too quickly. He said he's lost more money to botched migrations than to hacks. In other words, a response intended to improve security can introduce its own operational risk if funds are moved incorrectly.
That warning adds a practical complication to the proposed precautions. A migration is not automatically safe just because it aims to reduce exposure. Buterin's experience, as described in the source, is a reason to weigh the risks of changing wallet arrangements alongside the theoretical threat to signatures.
For users and the blockchain industry, the debate therefore has two parts: how to limit the information available to a potential attacker, and how to avoid mistakes during any migration. The source does not say that users should rush to move funds; it presents Drake's recommendation alongside Buterin's explicit caution about acting too fast.
Long-term security may need a different foundation
Looking beyond an immediate response, Buterin argues for an architecture that relies as much as possible on hash functions alone. The reason he gives is that even lattice-based schemes, often considered quantum-safe, could be weakened by advances in AI.
This points to a broader design question: how much should wallet security depend on signature methods that could be affected by future mathematical progress? Buterin's argument favors relying more heavily on hash functions. It is a long-term direction, not a report that a replacement has already been implemented.
For now, the reported facts support caution rather than certainty. Researchers are discussing a possible future risk, Drake has urged the industry to prepare, and Buterin has stressed the hazards of hurried migrations. The current ECDSA signature scheme has not been broken in practice, so the debate is about preparedness and the choices that could shape wallet security over time.