What it is. Avalanche splits into three connected chains plus optional custom chains, and the team has talked about quantum-safe math but put none of it into the live network.
What we found. A custom Avalanche chain could in theory turn on quantum-safe signing on its own, yet not one has done so, no quantum-safe code runs anywhere, and the very signatures validators use to confirm cross-chain transfers have no replacement plan at all.
Why it matters. A future quantum computer could forge those signatures and unlock funds, and on the chain that runs ordinary smart contracts your account is exposed the moment you make your first transaction, with the safe-by date already past the point experts treat as comfortable.
ACP-77 + HyperSDK permit a custom-VM Avalanche L1 to ship a PQ signature scheme without Primary Network coordination, a parallel-deployment capability that no documented Avalanche L1 has used. The BFT aggregation-path declaration (4f=0) is the consensus-critical gap: ACP-77's authorisation flow is BLS12-381-aggregation-bound at every step with no PQ replacement specified.
Summary
Avalanche scores QRI 23, Band 3 Planning, Migration Stage 0. Primary Network runs ECDSA secp256k1 (P-/X-/C-Chain account signatures, validator block-proposal signing under Parlia), BLS12-381 with KOSK (validator multi-sigs added in Banff 2022, used by Avalanche Warp Messaging and ACP-77 P-Chain authorisation), Keccak-256 (C-Chain), SHA-256 (P-/X-Chain), and RIPEMD-160 (NodeID and address derivation). Every signature, KEM-equivalent, and aggregation primitive in active production use is quantum-vulnerable. ACP-77 (Etna / Avalanche9000, Dec 2024) introduced an externalised validator-manager pattern; HyperSDK custom-VM L1s can in principle ship PQ schemes without Primary Network hard fork. Co-founder Sirer's Dec 2024 statement that Avalanche has lattice implementations developed but not deployed because lattice signatures are roughly 10× elliptic-curve size is the only foundation-tier signal of internal PQ work. AIP-QR-001 (PR #278, Feb 2026) is third-party-authored, not Ava-Labs-endorsed. mainnet-traffic cap binds at 5a=0%, Architecture-Execution Gap is 45, supply-chain weak-link cap fires across all 4 vendor tiles. Mining concentration on AWS (>33% of stake) is a real Dim 7 risk signal.
What the gates say
- Gate 1a, Hybrid signature: FAIL , no documented hybrid signature composition AND or OR with commit-to-hash at Primary Network level; the third-party AIP-QR-001 forum proposal is not Ava-Labs-endorsed
- Gate 1a, Hybrid KEM: FAIL , validator gossip TLS uses classical X25519 / ECDHE; no documented hybrid PQ KEM
- Gate 1b, Commit-to-hash: COND , no OR-composition deployed
- Gate 2, Evidence reconstruction: PASS , every sub-score reconstructible from public artifacts within 48 hours
- Gate 3, Primitive naming: PASS , every sub-score names specific primitives
Burn-vs-rescue policy on file
Declared option f, Undeclared. No Ava Labs / Avalanche Foundation policy on legacy-key handling under quantum-attack scenarios. Co-founder December 2024 statements are about Bitcoin P2PK, not Avalanche policy.
Seven dimensions
Each dimension scores 0–100 internally; the weighted roll-up produces the QRI.
1 Cryptographic Exposure weight 15% 29 / 100
Avalanche publishes a cryptographic-primitives standards page that names every active primitive at the Primary Network layer. Banff (AvalancheGo v1.9.0, 2022) added BLS12-381 multi-signatures with public-key aggregation.
ECDSA secp256k1 (P-/X-/C-Chain account signatures, validator block-proposal signing under Parlia) · BLS12-381 with KOSK (Primary Network validator multi-sigs added Banff 2022, AWM / ICM, ACP-77 messages) · Keccak-256 (C-Chain hashing) · SHA-256 (P-/X-Chain hashing) · RIPEMD-160 (NodeID and address derivation) · TLS with self-signed leaf certificates (validator transport) Every signature, KEM-equivalent, and aggregation primitive in active production use is quantum-vulnerable.
ECDSA-secp256k1→ Shor-break-via-DL-without-pairingsBLS12-381→ Shor-break-via-pairingsTLS validator gossip (X25519 / ECDHE / RSA-CA-equivalent self-signed)→ Shor-break-via-DL-without-pairings + Shor-break-via-RSA where applicableSHA-256 / SHA3-Keccak-256→ Grover-weaken-128bitRIPEMD-160→ Grover-weaken-80bit
0 PQ-safe families deployed. No hash-based, lattice, or code-based primitive in avalanchego cryptography or in any Primary Network chain. Subnets / Avalanche L1s under ACP-77 may select their own VM and primitive set, but no shipped subnet has been documented to deploy a PQ family.
No PQ primitive deployed at Primary Network or in any documented subnet. No production primitive maps to FIPS 203/204/205 categories.
Production primitives use libsecp256k1 (ECDSA) and supranational/blst (BLS12-381 aggregation, the same library Ethereum and Aptos consensus rely on). AvalancheGo has external audit history (Halborn, OpenZeppelin Warp Messaging audit). No formal-verification artifact published for the cryptography modules.
2 Quantum Recovery Exposure weight 10% 26 / 100
C-Chain accounts (EVM model) reveal the public key on first signed transaction. P-Chain and X-Chain addresses follow Bitcoin-style hashed-pubkey derivation, so unspent UTXOs that have never been moved retain a hash barrier. Validator BLS public keys are publicly registered on the P-Chain by design, every validator's BLS pubkey is part of the canonical validator set used for Warp / Interchain Messaging verification.
Mainnet genesis 21 September 2020 (~5.5 years operational by May 2026). Three-chain architecture means dormant value sits across two address models. Co-founder publicly stated in December 2024 that for Avalanche specifically, public keys are only revealed for a short time during a transaction, which applies to the X-/P-Chain UTXO model but not to C-Chain EVM accounts.
All historical ECDSA secp256k1 signatures across P-/X-/C-Chain since 2020 are forgeable post-Shor. All historical BLS12-381 multi-signature aggregates produced under AWM / ICM since the Banff upgrade (2022) are forgeable post-Shor; this includes ACP-77 SubnetToL1ConversionMessage, RegisterL1ValidatorMessage, L1ValidatorRegistrationMessage, and L1ValidatorWeightMessage types.
Validator-to-validator transport is TLS with self-signed leaf certificates (Avalanche-specific design). Underlying TLS handshake KEM is the standard classical X25519 / ECDHE option set; no documented hybrid PQ KEM deployment on validator gossip, RPC, or relayer channels.
3 Metadata, Anonymity & Confidentiality weight 13% 30 / 100
Pseudonymous transparent ledger across all three Primary Network chains. C-Chain is account-model EVM; X-Chain is UTXO-model with slightly higher unlinkability between outputs of the same wallet. No native shielded pool.
C-Chain RPC dominated by Ava Labs RPC, Infura, Alchemy, QuickNode, and Ankr. Validator hosting heavily concentrated on AWS, independent measurements report >33% of stake hosted on AWS (operational hosting Nakamoto coefficient ≈ 1) and >45% of validators using AWS overall.
Primary cross-chain routes are the Avalanche Bridge (originally federated; Intel SGX-based design), LayerZero, Wormhole, and Chainlink CCIP. All four are passively observable. Native Avalanche Interchain Messaging (BLS-aggregated) is observable to validators by design.
Standard transparent ledger model. Shor on secp256k1 exposes every historically-revealed C-Chain pubkey. Shor on BLS12-381 exposes every historical Warp / ICM aggregate signature, allowing full retrospective reconstruction of which validator set authorised which cross-subnet message at which P-Chain height.
4 Migration Architecture weight 10% 53 / 100
ACP-77 (Reinventing Subnets, Activated Dec 2024 via Etna / Avalanche9000) introduces an externalised validator-manager pattern; each Avalanche L1 sets a managerChainID and managerAddress. The HyperSDK custom-VM framework allows authors to register custom Auth modules. The Primary Network signature scheme cannot be swapped without a network upgrade ACP, there is no live precedent of swapping a Primary Network signature primitive.
C-Chain inherits Ethereum-equivalent account abstraction surface, ERC-4337 contracts deploy unchanged. HyperSDK custom-VM L1s can implement custom AA at the Auth-module level. There is no documented protocol-level key rotation primitive for P-Chain or X-Chain UTXO accounts. No PQ-active client-layer deployment.
Coordinated upgrades within 3 years: Banff (2022, BLS multi-sig added), Cortina, Durango (2024), Etna / Avalanche9000 (16 December 2024, ACP-77 activation). No contested forks; no ANC fork divergence.
Architectural support for hybrid composition exists: subnet / Avalanche L1 model and HyperSDK custom Auth modules permit AND-/OR-style multi-scheme transaction authorisation at the application layer. No ACP currently specifies a Primary Network hybrid (ECDSA AND PQ, or ECDSA OR PQ with commit-to-hash) for P-/X-/C-Chain accounts. The third-party AIP-QR-001 PR #278 sketches a hybrid validator signing model with C-Chain precompiles for Falcon and Dilithium verification, but draft, no merged code, no Ava Labs endorsement.
No stateful hash-based scheme deployed or proposed for Primary Network or any documented subnet. Default 15 applies.
Avalanche uses BLS12-381 multi-signature aggregation with public-key aggregation in two consensus-critical surfaces: Primary Network validator BLS keys aggregated under AWM / ICM, and ACP-77 P-Chain authorisation of Avalanche L1 validator-set state transitions where 67% of sourceChainID weight must participate in the aggregate BLS signature. No published spec, testnet, or mainnet pilot for a PQ aggregation-path replacement.
5 Deployment Execution weight 22% 8 / 100
0% of Primary Network signing traffic uses PQ primitives. No P-Chain, X-Chain, or C-Chain account type is callable on mainnet under a PQ scheme. No documented Avalanche L1 / subnet has shipped a PQ primitive in production.
No PQ signature implementation merged into avalanchego. Cryptography directory shows secp256k1, BLS12-381, Ed25519 (HyperSDK auth), SHA-256 / Keccak-256 / RIPEMD-160 modules; no slh_dsa, ml_dsa, ml_kem, falcon, or dilithium module.
No validator runs a PQ consensus key. ~1,400+ Primary Network validators use BLS12-381 keys for ICM / Warp aggregation.
VOIDED to 0 per v3.1 because 5a = 0. No Ava-Labs- or Avalanche-Foundation-published, enforcement-mechanism-backed dated PQ milestones. The third-party AIP-QR-001 forum proposal contains aspirational stages but no enforcement mechanism.
Announcements: co-founder December 2024 X / press cycle on Bitcoin P2PK quantum risk and statement that Avalanche has lattice cryptography implementations developed but not deployed; AIP-QR-001 forum thread (February 2026, third-party). Shipped: 0 mainnet PQ bytes signed, 0 ACPs activated. Not a sustained marketing posture.
No PQ signature footprint on mainnet. Co-founder December 2024 statement quantifies the design constraint: lattice signatures (Falcon-512 ~666 bytes, ML-DSA-65 ~3,309 bytes) are roughly an order of magnitude larger than secp256k1 ECDSA (~71 bytes). No published throughput-mitigation plan.
6 Supply Chain Vendor Readiness weight 22% 6 / 100
Top-3: Core Wallet (built by Ava Labs), MetaMask (via custom RPC for C-Chain), Ledger hardware. No published PQ roadmap on core.app, metamask.io, or ledger.com (PQ research / lattice prototyping is referenced by some hardware vendors but no concrete shipping commitment for Avalanche signing keys).
Top-3: Avalanche Bridge (Ava Labs / Intel SGX-based), LayerZero, Wormhole. None publishes a PQ roadmap for cross-chain signing or relayer transport keys. Avalanche Interchain Messaging itself is BLS12-381-based with no PQ alternative.
Top-3: Coinbase Custody, BitGo, Fireblocks. None publishes an Avalanche-specific PQ migration commitment. Fireblocks has publicly discussed MPC-PQ research; no production Avalanche signing key under PQ MPC.
Top RPC: Ava Labs RPC, Infura, Alchemy, QuickNode, Ankr. None publishes a hybrid-PQ TLS termination or PQ-attestation roadmap. AWS Nitro / KMS PQ migration is the relevant supplier for AWS-hosted validators, but no Avalanche-specific commitment. The Avalanche Bridge originally relied on Intel SGX TEE attestation; SGX attestation chains are classical-ECDSA-rooted.
7 Governance & Coordination weight 8% 46 / 100
Active Primary Network validators ~1,400-1,500 (~1,434 reported in 2025 ecosystem coverage; ~248M AVAX staked). Geographic distribution: independent measurements report US ~32%, Germany ~15%, Japan ~10% of stake. Hosting concentration is the weakness, >33% of stake on AWS yields an operational hosting Nakamoto coefficient = 1. Single AvalancheGo client.
Coordinated upgrades, Banff, Cortina, Durango, Etna / Avalanche9000 (December 2024, ACP-77 activation), executed without contested forks. ACP-77 alone restructured the relationship between L1s, validators, and the P-Chain.
Ava Labs (CEO Emin Gün Sirer; co-founders Kevin Sekniqi, Maofan Ted Yin) is the principal AvalancheGo client developer; the Avalanche Foundation maintains the ACP repository and ratification process. No named PQ-migration working group, lead, or published mandate.
No documented coordinated cryptographic-primitive change executed under live attack. No quantum-tripwire incident response. Standard upgrade cadence is non-adversarial.
No published canary, honeypot, rate-limited spending rule, or cryptographic tripwire embedded in consensus targeting quantum-attack detection.
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 range 2030–2046, upper portion Outside risk window vs Z25; Crisis Zone vs Z10 for slower X scenarios
Z-compliance
Outside compliance window for 2035 disallowance under realistic Y
Source-disagreement disclosure
v3.1 requires every chain card to publish material divergences among authoritative sources, plus the delta-QRI under alternative weighting.
Independent ecosystem trackers (avascan.info, Messari) vs Avalanche Foundation press cite 1,400-1,700 ranges across 2024-2026; we use 1,434 (2025) as a current-cycle midpoint.
Messari and Nakaflow report >33% AWS stake share; Avalanche Foundation does not publish a hosting-distribution figure.
Delta-QRI under alternative weighting
Under heavier Dim 4 weight (architectural sovereignty model), Avalanche's score rises modestly; under heavier Dim 5 / Dim 6 weights (deployed PQ + supplier readiness), score falls to high-teens. Range ≈ 19-28.
Announcement-to-shipped ratio
Announced: 2. Shipped: 0. Ratio: 2.
Tag: none
Peers in the L1 profile
9 chains closest to Avalanche by Stage then QRI.