On October 7, 2026, OpenAI released 722 mathematical manuscripts spanning number theory, geometry, and theoretical computer science. Within hours, Ethereum Foundation researcher Justin Drake called for the crypto industry to prepare "bunker mode" — a controlled migration of assets to fresh addres...
"There is now reason to prepare for the possibility that ECDSA may fail. The worst case is months, not years." — Justin Drake, Ethereum Foundation Researcher
On October 7, 2026, OpenAI released 722 mathematical manuscripts spanning number theory, geometry, and theoretical computer science. Within hours, Ethereum Foundation researcher Justin Drake called for the crypto industry to prepare "bunker mode" — a controlled migration of assets to fresh addresses with unexposed public keys. Ethereum co-founder Vitalik Buterin followed, warning that AI-accelerated mathematics has a "good chance" of weakening lattice-based cryptography, including the NIST-approved ML-DSA standard, within two years.
The warnings landed one day after Europol's EC3 unit flagged 6.9 million BTC ($586 billion) and 55–60% of all ETH as quantum-exposed through on-chain public key leakage. Three major Layer 1 networks — Bitcoin, Ethereum, and Solana — now face overlapping AI and quantum threats to their signature schemes. Each has chosen a different migration architecture, timeline, and set of trade-offs. This report compares them.
The crypto industry has long treated quantum computing as the primary cryptographic threat. Drake's October 7 statement reframed the risk calculus: AI-driven mathematical breakthroughs may crack ECDSA — the elliptic curve signature scheme securing Bitcoin, Ethereum, and Solana transactions — before quantum hardware reaches sufficient scale.
Drake's worst-case scenario: a private-key recovery attack executable in approximately one week using a large GPU cluster. This contrasts with the quantum timeline, where most estimates place a cryptographically relevant quantum computer at 2030 or later.
Buterin's concern extends further. He warned that lattice-based cryptography — the mathematical foundation underlying NIST's post-quantum standards ML-KEM (FIPS 203) and ML-DSA (FIPS 204), finalized August 13, 2024 — could itself be weakened by AI within two years. His recommendation: shift toward pure hash-based signature schemes such as WOTS and SPHINCS+, which rely on simpler mathematical assumptions.
The implication is significant. Networks planning to migrate from ECDSA to lattice-based post-quantum schemes may need to migrate twice — first from ECDSA to lattice, then from lattice to hash-based — if Buterin's assessment proves correct.
Exposure: Approximately 6.7–6.9 million BTC (34% of circulating supply) sit in addresses with public keys exposed on-chain, according to Glassnode and Europol EC3 data. Of these, 1.92 million BTC (9.6% of issued supply) carry structural exposure from legacy P2PK address formats, while 4.12 million BTC (20.6%) have operational exposure from transaction-signing activity.
Roughly 1.7 million BTC in P2PK outputs date to 2009–2010, including an estimated 1.1 million BTC attributed to Satoshi Nakamoto across approximately 22,000 outputs. These coins cannot be moved without their creator's private keys.
Migration Plan: BIP-361, published in April 2026 by Jameson Lopp and five co-authors, proposes a three-phase sunset of legacy ECDSA and Schnorr signatures:
Trade-off: Bitcoin's approach forces a binary choice — freeze vulnerable coins or accept the risk of theft. Phase B would render 6.7+ million BTC unspendable until Phase C recovery mechanisms are developed and activated. The proposal requires broad social consensus. As of October 2026, BIP-361 remains a draft.
Signature Scheme: BIP-360 specifies a new P2MR (Pay to Merkle Root) output type. The specific post-quantum signature algorithm has not been finalized. Candidates include hash-based schemes and lattice-based approaches.
Exposure: Europol's EC3 estimated 55–60% of all ETH sits in addresses with public keys exposed on-chain, a higher proportional exposure than Bitcoin.
Migration Plan: The Ethereum Foundation formed its post-quantum security team in January 2026 and launched pq.ethereum.org as a coordination hub. Over 10 client teams participate in weekly interoperability devnets. The Foundation allocated a $1 million "Poseidon Prize" for research into hash-based cryptographic primitives.
Buterin's roadmap, published in February 2026, identifies four vulnerable cryptographic layers:
The key architectural advantage: EIP-8141 introduces native account abstraction, allowing individual accounts to choose their own signature verification logic. Users can migrate to quantum-safe signatures independently, without waiting for a protocol-wide hard fork. EIP-8141 is under consideration for the Hegotá upgrade.
Trade-off: Ethereum's approach is more flexible but slower. Core post-quantum infrastructure milestones target completion by approximately 2029, according to the "Lean Ethereum" roadmap. The three-year gap between now and full deployment represents a window of vulnerability. Additionally, Buterin's own warning about lattice fragility complicates the choice of replacement algorithm — the Foundation is now accelerating work on hash-based signatures rather than relying solely on NIST-approved lattice standards.
Signature Scheme: Moving toward hash-based (WOTS, SPHINCS+) rather than lattice-based (ML-DSA). Buterin explicitly favored this approach on October 7.
Exposure: Solana uses Ed25519 (a variant of EdDSA on Curve25519), which carries the same quantum vulnerability class as ECDSA. Specific exposure statistics for Solana's address base were not available in reviewed sources.
Migration Plan: In April 2026, Solana's two core client teams — Anza and Jump Crypto's Firedancer — independently converged on the Falcon signature scheme as their post-quantum solution. The roadmap specifies a phased transition:
Trade-off: Performance degradation is the primary constraint. Quantum-safe Falcon signatures are 20–40x larger than Ed25519 signatures. Internal testing showed a Falcon-enabled Solana validator running approximately 90% slower than current production. For a network that markets sub-second finality and high throughput, this represents an existential product trade-off.
Solana partnered with Project Eleven in late 2025 for quantum threat assessment. The network's approach prioritizes readiness over immediacy — migration mechanisms are designed but not activated.
Signature Scheme: Falcon (a lattice-based scheme). Notably, this is the category Buterin flagged as potentially AI-vulnerable within two years.
The migration challenge compounds for institutional custodians. Banks, exchanges, and asset managers hold assets across multiple chains, each adopting different post-quantum standards on different timelines.
On October 8, 2026, Project Eleven and Quantus announced plans to integrate Quantus Network with Strongpoint, Project Eleven's institutional custody platform. The integration, targeted for Q1 2027, addresses this fragmentation through a "crypto-agile" architecture: authentication, policies, approvals, key management, and audit operate through a common control layer, separated from the cryptography of the underlying network.
Drake singled out several major custodians — Binance, Bitbank, Robinhood, Bitfinex, and Tether — as institutions that should strengthen cold-storage practices. Centralized trading platforms hold roughly 1.66 million BTC of the network's quantum-exposed supply, representing the single largest operational hotspot.
The CoinDesk analysis published October 8 noted that "bunker mode is a far greater challenge for institutions than individual crypto holders." The core difficulty: each blockchain may adopt a different quantum-resistant standard (hash-based for Ethereum, Falcon for Solana, TBD for Bitcoin), forcing custodians to support multiple cryptographic backends simultaneously while maintaining existing compliance and audit controls.
| Dimension | Bitcoin | Ethereum | Solana | |---|---|---|---| | Current Signature | ECDSA / Schnorr | ECDSA | Ed25519 | | Exposed Supply | 6.7–6.9M BTC (34%) | 55–60% of ETH | Not quantified | | Migration Proposal | BIP-361 (3-phase sunset) | EIP-8141 (account abstraction) | Phased wallet migration | | Target PQ Scheme | TBD (P2MR output type) | Hash-based (WOTS, SPHINCS+) | Falcon (lattice-based) | | Performance Impact | Minimal (UTXO model) | Moderate (larger signatures) | Severe (~90% throughput loss) | | Legacy Coin Handling | Freeze, then ZK recovery | Per-account migration | Migrate when threat materializes | | Full PQ Target Date | ~8 years post-activation | ~2029 | Not specified | | Status | Draft BIP | Active research, 10+ client teams | Design complete, not activated | | AI-Lattice Risk | N/A (scheme TBD) | Mitigated (choosing hash-based) | Exposed (Falcon is lattice-based) |
The threat model has shifted. Drake and Buterin's October 7 statements reframe the timeline from "quantum in 2030+" to "AI-assisted ECDSA break possibly within months." This is a tail-risk scenario, not a base case, but it has accelerated planning across all three networks.
No chain is post-quantum today. Bitcoin, Ethereum, and Solana all rely on quantum-vulnerable signature schemes in production. Migration plans exist at various stages of maturity, but none have activated network-wide changes.
Ethereum's account abstraction approach offers the most flexible migration path, allowing individual accounts to upgrade without protocol-wide coordination. However, full implementation targets 2029.
Bitcoin's BIP-361 is the most aggressive proposal but requires freezing potentially millions of BTC in legacy addresses — a socially and economically contentious step that lacks consensus.
Solana faces a unique performance trade-off. Its chosen scheme (Falcon) delivers 20–40x signature bloat and approximately 90% throughput loss in testing. Additionally, Falcon's lattice-based construction falls within the category Buterin flagged as AI-vulnerable.
Institutional custodians face the worst complexity. Multi-chain portfolios must support divergent cryptographic backends across Bitcoin, Ethereum, Solana, and other networks — each migrating on its own timeline and to its own standard.
The AI threat to lattice cryptography may force a second migration. If Buterin's two-year warning proves correct, networks that migrate to lattice-based schemes (Solana's Falcon, potentially Bitcoin's TBD scheme) may need to re-migrate to hash-based alternatives.
The convergence of AI mathematical breakthroughs and quantum computing progress has compressed the timeline for post-quantum migration across major blockchain networks. As of October 8, 2026, no production blockchain has activated quantum-resistant signatures. Bitcoin, Ethereum, and Solana have each proposed distinct migration architectures reflecting their design philosophies: Bitcoin's conservative freeze-and-recover approach, Ethereum's modular account-abstraction path, and Solana's performance-constrained phased rollout.
The economic stakes are quantifiable: $586 billion in quantum-exposed BTC alone, plus majority exposure across Ethereum's supply. The institutional custody layer adds operational complexity that individual migration plans do not address.
What remains unresolved is whether lattice-based cryptography — the foundation of NIST's approved post-quantum standards and Solana's chosen scheme — will itself prove durable against AI-accelerated mathematical research. If it does not, the industry faces not one migration but two.