What it is. Arbitrum One is the busy Ethereum companion network that holds more user money than any rival of its kind, and it is rare in that developers can add brand-new security checks here without waiting for the whole Ethereum network to upgrade.
What we found. That same head start sits completely unused for quantum defense: nobody has switched on a tougher way to sign transactions, nobody has run a trial, no date is on the calendar, and no one inside the company that runs it owns the job.
Why it matters. The chain best placed to act early is the one moving last, so anyone parking large balances here is trusting that future quantum machines stay too weak to fake account approvals, with no public sign that the team will move before that day arrives.
Arbitrum One has every architectural surface a credible PQ migration would need (Stylus WASM precompiles, EIP-7702 EOA delegation, EIP-2537 BLS precompiles, ERC-4337 since 2023) and zero PQ-specific intent: no AIP, no public spec, no testnet hybrid, no foundation position, no named PQ lead. The L2 most architecturally ready to migrate that has documented-publicly the least about migrating.
Summary
Arbitrum One scores QRI 26, Band 2 Acknowledged, Migration Stage 0. As a rollup-L2 settling on Ethereum, it inherits user-layer ECDSA secp256k1 / Keccak-256 and posts batches as Ethereum calldata or EIP-4844 blobs signed by the batch poster. ArbOS 51 Dia (activated 2026-01-08) enabled EIP-2537 precompiles for BLS12-381 curve operations. ArbOS 40 Callisto (activated 2025-06-17) enabled EIP-7702 EOA delegation. Stylus WASM contracts let developers deploy custom precompiles for any cryptographic primitive without an L1 hard-fork, genuinely novel agility for a rollup. BoLD permissionless validation activated 2025-02-12; Stage 1 decentralization with 12-member Security Council. Arbitrum is not a member of the Optimism Superchain, Coinbase's separation announcement (2026-02-18) was for Base, not Arbitrum. There is no announced Arbitrum-specific PQ migration roadmap; trailing-12-month PQC announcement count from Offchain Labs / Arbitrum Foundation is zero. mainnet-traffic cap binds at 5a=0%, Architecture-Execution Gap is 53.5, supply-chain weak-link cap fires across 4 of 4 vendor tiles. The agility surface is live; migration intent is absent.
What the gates say
- Gate 1a, Hybrid signature: FAIL , no documented hybrid signature composition AND or OR at user signing or sequencer signing layer
- Gate 1a, Hybrid KEM: FAIL , no documented hybrid KEM at sequencer-RPC TLS or DAS gossip
- Gate 1b, Commit-to-hash: COND , no OR-composition declared
- Gate 2, Evidence reconstruction: PASS , every non-voided sub-score has 3+ public artifacts
- Gate 3, Primitive naming: PASS , every named primitive specified
Burn-vs-rescue policy on file
Declared option f, Undeclared. No Arbitrum Foundation or Offchain Labs position on what happens to dormant L2 funds at user EOAs in a quantum scenario. Rollup-L2-specific consideration: dormant L2 funds depend on the canonical bridge contract's L1 ownership; if Ethereum L1 freezes vulnerable EOAs, the L2 mirror state inherits that freeze automatically. If a user has only ever interacted at L2, there is no Ethereum L1 fingerprint to freeze. No published position.
Seven dimensions
Each dimension scores 0–100 internally; the weighted roll-up produces the QRI.
1 Cryptographic Exposure weight 12% 34 / 100
Arbitrum One inherits Ethereum's primitives at the user layer and posts batches as Ethereum calldata or EIP-4844 blobs. ArbOS 51 Dia enabled EIP-2537 BLS12-381 precompiles. BoLD's fraud-proof game uses deterministic Merkle proofs over Keccak-256 commitments to WAVM execution traces. Stylus contracts execute compiled WASM and can implement custom cryptographic verifiers, but no PQC primitive shipped in any audited Stylus precompile.
ECDSA secp256k1 (EOA signatures) · Keccak-256 (hashing/Merkle tries) · BLS12-381 (EIP-2537 precompiles, post-ArbOS 51 Dia) · AES-128-GCM (TLS, sequencer feed) · X25519 (TLS handshake) · SHA-256 (Ethereum settlement, Merkle proofs) ECDSA secp256k1→ Shor-break-via-DL-without-pairingsKeccak-256→ Grover-weaken (256→128-bit preimage)BLS12-381 (post-ArbOS 51 precompile, opt-in for contracts and underlying ETH consensus signatures Arbitrum settles into)→ Shor-break-via-pairingsSHA-256→ Grover-weakenAES-128-GCM→ Grover-weaken to 64-bit security
0 PQ families deployed. Arbitrum One has zero post-quantum cryptography in active production use. No lattice, hash-based, code-based, or isogeny family has shipped.
No NIST FIPS 203/204/205 PQC primitive deployed by Arbitrum or Offchain Labs. No public mapping exists.
BoLD reference implementation audited by Trail of Bits (multiple reports 2024-2025) but no machine-checked formal proofs published. Relies on go-ethereum's crypto/secp256k1 and standard libsecp256k1 wrappers (industry-standard but not Formosa/Jasmin-grade). Tier 1 (classical ECC + Keccak-256), Tier 2 for BLAKE-style internals.
2 Quantum Recovery Exposure weight 8% 19 / 100
Arbitrum One holds approximately $15.2B–$16.8B TVL (L2BEAT, DefiLlama 2026 Q1), the largest of any L2. EOAs use ECDSA secp256k1; the public key is revealed on first transaction. All TVL sits behind keys whose pubkeys are or can be derived from observed signatures.
Arbitrum One has been live since 2021-08-31 (~56 months at evaluation date). Significant unmoved-balance addresses exist (bridge contracts, treasuries, dormant LP positions). Ethereum-format addresses do not directly reveal the pubkey, but any prior outbound tx already exposed the pubkey on Ethereum or Arbitrum.
Sequencer batch-poster ECDSA signatures and Ethereum L1 state-root commitments are forgeable post-Shor against historical chain history. No retro-signing or post-fact PQ wrapping mechanism exists.
Sequencer feed and public RPC endpoints served over standard TLS 1.3 (typically X25519/P-256 ECDHE + AES-GCM). No documented hybrid PQ KEM in use on the public sequencer feed. Validator-to-DAS gossip on Nova uses BLS-key authentication for committee membership but session encryption defaults to TLS.
3 Metadata, Anonymity & Confidentiality weight 8% 17 / 100
Pseudonymous transparent EVM ledger; full tx graph public.
Tx ordering controlled by a single Offchain Labs sequencer (Nakamoto coefficient = 1 for sequencing). Public RPC concentrated among Alchemy/Infura/QuickNode. No public mempool, txs go directly to the sequencer, then publish in batches; this reduces front-running surface but increases observability of pre-confirmation order to the operator.
Canonical Arbitrum Bridge produces direct on-chain linkage between Ethereum L1 deposits and L2 receipts. Third-party bridges (LayerZero, Hop, Across) make the L1↔L2 hop traceable through public observers.
No shielded pool on Arbitrum One. ECDSA Shor-breaks the limited privacy that exists from address pseudonymity once linked to identity off-chain.
No on-chain mixer integrated into Arbitrum One protocol.
4 Migration Architecture weight 15% 67 / 100
Arbitrum has live, production-shipped crypto-agility surfaces. EIP-7702 (EOA delegation) is live since ArbOS 40 Callisto (activated 2025-06-17 on Arbitrum One and Nova). EIP-2537 BLS12-381 precompiles are fully enabled in ArbOS 51 Dia (activated 2026-01-08). Stylus (live since 2024) lets developers deploy WASM-based custom precompiles for any cryptographic primitive without an L1 hard-fork, genuinely novel agility for a rollup. ArbOS upgrades go through DAO + Security Council governance with established cadence. No protocol-level algorithm switch for the user-EOA signature scheme has shipped, however, so the agility is presently latent.
ERC-4337 supported on Arbitrum One since 2023; EIP-7702 live since 2025-06-17 enabling EOAs to delegate to smart-contract code without changing address. Client-layer PQ migration is architecturally tractable: a user could opt into a smart-account that verifies an ML-DSA or SLH-DSA signature inside contract code or via a Stylus precompile. Arbitrum is NOT a member of the Optimism Superchain. No Arbitrum-specific client-layer PQ migration formally documented or deployed.
Coordinated upgrade record over 36 months: Nitro migration (2022), Stylus mainnet (2024), BoLD permissionless validation (activated 2025-02-12), ArbOS 31/32 (2024), ArbOS 40 Callisto (2025-06-17), ArbOS 51 Dia (2026-01-08). Each upgrade went through DAO vote + Security Council approval + multi-week timelock. No contested or stalled forks.
Stylus + EIP-7702 + smart-contract wallets make a hybrid (classical ECDSA + PQ ML-DSA) signature path architecturally constructible. No such hybrid has been spec'd or shipped by Arbitrum or Offchain Labs. The architectural primitives exist; the migration document does not.
Arbitrum has no stateful hash scheme (XMSS/LMS/leanXMSS) at the protocol level; default-pass per v3.1.0 rule.
N/A. Arbitrum One is a rollup with a single sequencer for transaction ordering and a permissionless fraud-proof game (BoLD) for state validity. There is no BFT consensus with BLS signature aggregation at the Arbitrum One layer; consensus is inherited from Ethereum L1.
5 Deployment Execution weight 22% 14 / 100
Zero PQ signatures observed in the Arbitrum One transaction stream. No PQ KEM in TLS termination documented for Offchain Labs sequencer endpoints. No Stylus precompile shipped that exposes ML-DSA/ML-KEM/SLH-DSA.
A grep of the Nitro reference client (github.com/OffchainLabs/nitro) yields no ML-DSA, ML-KEM, SLH-DSA, Falcon, XMSS, or SPHINCS+ implementation. The crypto package uses go-ethereum primitives.
The sequencer is centralized at Offchain Labs (single signer, secp256k1). BoLD's permissionless validators bond using ETH and post Merkle-proof commitments under standard Ethereum-format keys; no PQ key adoption among validators.
VOIDED to 0 per v3.1 rule (5a = 0). No PQC-specific dated milestones published by Offchain Labs, the Arbitrum Foundation, or any Arbitrum DAO AIP.
Trailing-12-month PQC announcement count from Offchain Labs / Arbitrum Foundation channels: 0. Shipped PQ on mainnet: 0. Ratio: 0/0 (treated as 1.0, no washing). Arbitrum has not over-promised; it has not promised at all. The cleanest possible washing posture but also the absence of a roadmap.
No PQ signature deployed; no mainnet bytes-per-block multiplier observable.
6 Supply Chain Vendor Readiness weight 25% 6 / 100
Top-3: MetaMask, Rabby, Safe (smart-account on Arbitrum). Hardware: Ledger, Trezor. Trezor Safe 7 (released 2025) ships ML-DSA-44 device attestation + SLH-DSA-SHA2-128s bootloader signature verification, the only top-3 wallet vendor with shipped PQ in production, and it is for device attestation/firmware integrity, not transaction signing. ERC-4337 / Safe modules CAN be extended with a PQ verifier (e.g., a Stylus contract), but no production deployment exists.
Top-3: canonical Arbitrum Bridge, LayerZero, Across. Canonical Arbitrum Bridge inherits ECDSA + Keccak-256 + (where applicable) BLS12-381 from Ethereum L1 settlement; no PQ verification path. LayerZero has stated quantum durability is on the long-term agenda but no shipped PQ DVN configuration. Tile is the chain's largest concrete PQ exposure.
Top-3: Coinbase Custody, BitGo, Fireblocks. All three support Arbitrum-native USDC, ETH, ARB. None have a publicly shipped PQ key migration on Arbitrum mainnet. Industry trade press describes custodians as piloting quantum-resistant key migration roadmaps aligned with NIST PQC standards, pilot, not shipped.
Top-3 RPC: Alchemy, Infura, QuickNode → no public PQ TLS termination roadmap. HSM: AWS KMS shipped ML-DSA support 2025-06 (cloud HSM signing for FIPS 204), Thales/YubiHSM pilots. TEE attestation chains classical. AWS KMS ML-DSA is the only concrete-deployed PQ infra primitive; it is generic, not Arbitrum-bound.
7 Governance & Coordination weight 10% 59 / 100
Sequencing centralized at Offchain Labs (NC=1 for ordering). BoLD permissionless validation live on mainnet since 2025-02-12 lets any party bond and post fraud proofs, material decentralization improvement at the validation layer, but ordering remains a single point of trust.
36-month track record of coordinated upgrades through DAO + 12-member Security Council. ArbOS releases Callisto (June 2025) and Dia (January 2026) executed on schedule. Sequencer outages (Sep 2023, Dec 2023) handled within hours with public RCAs.
Offchain Labs (Co-founders Steven Goldfeder, Ed Felten, Harry Kalodner, Princeton-origin) is the named protocol-development entity. Arbitrum Foundation handles governance and treasury. Both have public mandates and disclosed boards. No designated PQ lead within either entity.
Sequencer outage incidents recovered without contested forks. No precedent of a coordinated cryptographic-primitive change under active attacker pressure. BoLD permissionless transition (Feb 2025) succeeded against the threat model of malicious validators rather than active in-flight adversary.
No quantum canary (community honeypot, rate-limit, cryptographic tripwire, or automated response) embedded in Arbitrum One protocol.
X + Y vs Z, when does the math turn against you?
v3.1 demotes the X+Y vs Z timing test to a secondary signal, the headline output is Migration Stage. The timing test still answers the question: can this chain finish migrating before the threat lands?
Verdict
X+Y > 2035 in all reasonable scenarios, Outside risk window vs Z25; Crisis Zone vs Z10
Z-compliance
Outside compliance window for any jurisdiction with hard-stop 2030 (NIST 2030 deprecation; CNSA 2.0 2030; EU NIS2)
Source-disagreement disclosure
v3.1 requires every chain card to publish material divergences among authoritative sources, plus the delta-QRI under alternative weighting.
L2BEAT and DefiLlama report TVL figures within a $15.2B-$16.8B band depending on date and inclusion rules (DeFi-only vs total bridged value); chain card uses the bridged-value figure consistent with L2BEAT Total Value Secured metric.
One alternative-weighting view places stronger emphasis on supply-chain readiness (lifting Dim 6 share toward 30%+) and explicitly recommends 1-of-2 hybrid signing as the canonical L2 user-layer migration path.
Delta-QRI under alternative weighting
Estimated −3 (i.e., QRI ≈ 23 if supply-chain weighted at 30%).
Announcement-to-shipped ratio
Announced: 0. Shipped: 0. Ratio: 0.
Tag: none
Peers in the rollup-L2 profile
9 chains closest to Arbitrum One by Stage then QRI.