Dragonfly Partner Rejects “Bunker Mode,” Urges Proactive Blockchain Security

Ethereum researcher Justin Drake has renewed urgency around the possibility that advances in AI—and, later, quantum computing—could undermine the cryptographic signatures used to secure crypto wallets. In comments posted this week, Drake suggested that the elliptic curve digital signature algorithm (ECDSA) could be broken on a timeframe measured in months rather than years, citing recent progress highlighted in an OpenAI report about AI in mathematics.

Dragonfly managing partner Haseeb Qureshi pushed back on a narrow “move to new addresses” approach, arguing that relocating funds to fresh public keys would not be sufficient protection if the wider network ecosystem remains vulnerable. Instead, Qureshi called for network-level safeguards, proposing a “Cryptographic Recovery Mode” designed to help validators recover funds even if signatures are compromised.

Key takeaways

• Justin Drake warned ECDSA could fail “in the worst case” within months, not years, based on rapid AI progress in mathematics.
• Dragonfly’s Haseeb Qureshi criticized “bunker mode” as an investor-focused fix that won’t stop mass compromise if others remain exposed.
• Glassnode data cited by Qureshi’s discussion suggests 6.26 million BTC may be vulnerable, including 4.33 million linked to address reuse.
• Qureshi proposed a “Cryptographic Recovery Mode” that would let validators force recovery through hash-based backup signatures.
• A large share of exposed Bitcoin sits with exchanges, with Glassnode noting nearly 1.8 million BTC on exchange balances.

Drake’s “months, not years” warning on ECDSA

Drake’s latest comments focus on the ECDSA mechanism that underpins wallet authorization in many parts of the crypto ecosystem. In a post on X, he argued it is “reasonable to brace” for ECDSA breaking earlier than many in the industry may have assumed, with the “worst case” timeframe measured in months rather than years. Drake pointed to a Tuesday OpenAI report discussing AI’s progress in mathematics as part of the rationale for taking the risk seriously.

Drake also recommended a practical user action: gradually moving funds to fresh wallets so that the associated public key is not exposed. His framing reflects a risk model where signature security can degrade faster than longer-term migration plans would typically account for—particularly if AI accelerates the relevant computations.

Vitalik Buterin indicated he agrees that the industry should take AI-accelerated math seriously, but he did not endorse rushing users to immediately relocate funds to new wallets. That divergence highlights a key tension in the current debate: what to do now versus what to do at the protocol and infrastructure level.

Qureshi rejects “bunker mode” as a self-contained solution

While Drake advocated for moving funds to avoid exposed public keys, Qureshi argued that the protective value of fresh addresses is far more limited than “bunker mode” framing suggests. In a Thursday X post, he characterized Drake’s warning as “cryptographic doomerism” and argued that migrating to new addresses only helps investors if they never need to move those holdings after the switch.

Qureshi’s core concern is network-wide. He warned that if cryptographic failures spread beyond a small set of “fresh” holders—ultimately enabling mass theft and selling—then even coins protected behind newly generated addresses could end up functionally worthless in practice. In his view, protecting individual keys does not solve the broader systemic problem if the underlying signature scheme becomes unreliable across the board.

Glassnode: millions of BTC exposed through reuse and key exposure

The discussion is not just theoretical. Qureshi pointed to data from Glassnode suggesting more than 31% of the Bitcoin supply may face some form of exposure, specifically citing 6.26 million BTC in vulnerable addresses.

According to Glassnode co-founder Rafael Schultze-Kraft, of the 6.26 million BTC, about 4.33 million BTC are exposed due to address reuse. In that scenario, moving funds to a fresh address could eliminate the reuse-related exposure. Separately, roughly 1.94 million BTC are exposed through address format, a different vector that is not solved simply by generating a new address.

Schultze-Kraft also highlighted where the exposed supply sits. Of the total exposed holdings, nearly 1.8 million BTC are held on cryptocurrency exchanges, with Glassnode noting that 57% of all exchange balances are currently exposed. This matters because exchanges are central to liquidation and custody flows; if signature compromise becomes a live threat, infrastructure that handles high transaction volumes could become a focal point for operational and risk-management decisions.

From migration to protocol recovery: Qureshi’s proposed “Cryptographic Recovery Mode”

Rather than relying only on end-user migration, Qureshi argued for proactive, network-level defenses against a scenario in which cryptographic signatures can be broken—whether due to AI-driven computation or later quantum capabilities. His proposal centers on a “Cryptographic Recovery Mode,” described as a hash-based backup signature plan that users could map to their addresses.

The mechanism, as Qureshi outlined it, would enable validators to force recovery if cryptographic signatures were compromised. The underlying idea is to build a path to regain control even when normal signature verification can no longer be trusted, shifting part of the security burden away from timing-dependent user actions.

For investors and builders, the distinction is important. Address migration is a mitigation strategy that depends on how quickly users can respond and whether their keys remain safe after movement. A recovery mode, by contrast, aims to reduce the all-or-nothing nature of “fresh address” protection by introducing a contingency structure at the validation layer.

Still, several uncertainties remain implicit in the debate. Drake’s recommendation is framed around reducing exposed public keys, while Qureshi’s counterproposal depends on the design and adoption of a recovery process that validators can execute. The market impact of either approach will likely depend not just on theoretical cryptography timelines, but on whether compatible standards can be implemented across wallets, exchanges, and validation infrastructure.

What to watch next as the timeline debate evolves

Drake’s “months, not years” warning has reframed the urgency around cryptographic resilience, but the most actionable next step for the industry may be clearer specifications of how wallets and networks should handle exposure—especially for the large portion of Bitcoin held by exchanges. Readers should watch for follow-up work that translates the recovery concept into implementable protocol changes and, separately, for guidance on how quickly practical migrations can realistically reduce exposure without creating operational risk.