What it is. Base is a network built on top of Ethereum, run by one publicly listed company, with about 4.7 billion dollars of user funds sitting in its apps.
What we found. No part of the live network protects those funds against a future quantum computer, and the only work under way sits on an experimental test network that holds no real money.
Why it matters. The company has put dates on protecting the customer funds it holds itself and none on the network where your own wallet sits, so the risk you carry personally is the part with no timetable.
Nothing post-quantum runs inside Base: EOAs and the batcher and proposer roles sign with ECDSA secp256k1, the passkey smart-account path uses ECDSA secp256r1 through the P256VERIFY precompile at 0x100 (present since Fjord via RIP-7212, mainnet 2024-07-10; re-priced to 6,900 gas at Azul per EIP-7951, mainnet 2026-05-28), and a code search of the base/base monorepo carrying base-consensus and base-reth-node returned zero hits on 2026-08-19 for ML-DSA, ML-KEM, SLH-DSA, Falcon and Kyber. Both hybrid gates fail at 0% mainnet post-quantum traffic, the corporate parent's 2026-07-23 roadmap commits Base only to a cryptographic inventory and to deferral to Ethereum's post-quantum roadmap, and the sole post-quantum-anticipating artifact is a reserved non-secp256k1 self-authenticator slot in Base's EIP-8130 implementation on the public Vibenet devnet, where a post-quantum verifier appears only as a code-comment example and the Cobalt upgrade announced for September 2026 ships the classical authenticator set (secp256k1, P-256, WebAuthn, delegate).
Summary
Base scores QRI 30, Band 3 Planning, Migration Stage 1. Settling on Ethereum, it carries ECDSA secp256k1 and Keccak-256 at the user layer plus a second signing path, WebAuthn on ECDSA secp256r1, verified by the P256VERIFY precompile at 0x100 present since Fjord (RIP-7212, mainnet 2024-07-10) and re-priced to 6,900 gas at Azul per EIP-7951 (mainnet 2026-05-28). Azul’s AggregateVerifier multi-proof path (AWS Nitro TEE provers attested under ECDSA P-384, permissionless ZK provers) is classical throughout. The corporate parent’s 2026-07-23 roadmap gives custody-side horizons (post-quantum rebuild of the key-management system covering approximately 99.9% of custodied assets, a post-quantum-capable signing pipeline within one year, lattice-based MPC signing over the following two to three years) and commits Base only to a cryptographic inventory. Base’s EIP-8130 implementation reserves a non-secp256k1 self-authenticator slot, a post-quantum verifier named as the example occupant, on the public Vibenet devnet only; Cobalt, announced for September 2026, ships a classical authenticator set. Mainnet post-quantum traffic is 0%, which voids 5d; both hybrid gates fail; the Architecture-Execution Gap is 64 and the recorded caps do not bind at 27. The one post-quantum primitive observable anywhere is ML-KEM-768 in the X25519MLKEM768 TLS key agreement at the operator’s CDN edge: transport, not protocol.
Forge. Forge-dominant: this chain secures value and operations with signatures, so the principal quantum risk is forgery of spends/attestations once Shor breaks the curve. There is no harvest-now component for forgery, because the public key alone enables it. Decrypt/HNDL applies only to transport/RPC confidentiality, where the operator's CDN edge already negotiates a hybrid ML-KEM-768 key agreement.
2 announced → 0 shipped on mainnet under a named primitive. announced-not-shipped, below deduction threshold. Two PQ-adjacent announcements touch Base in the trailing 12 months: the corporate parent's 2026-07-23 roadmap (names Base in a company-wide cryptographic-inventory commitment; names no Base primitive and sets no Base date) and Base's own 2026-07-17 native-account-abstraction announcement, which lists 'Quantum Ready' (key rotation and multiple authentication schemes 'allow users to upgrade to post quantum authentication') among EIP-8130's features while the canonical authenticator set contains no post-quantum scheme. Nothing post-quantum has shipped on Base mainnet. Neither announcement names a primitive, so neither is a press-release primitive claim in the strict sense of the 5e rule; with zero shipped the quotient is not numerically meaningful and is recorded at 1.0 with this flag rather than as an unbounded ratio. Re-check when the promised Base-specific plan publishes and when Cobalt activates..
What the gates say
- Gate 1a, Hybrid signature: FAIL , no documented hybrid signature composition AND or OR; the issuer's smart-account product's multi-owner architecture supports parallel verification of secp256k1 + secp256r1, both classical; the EIP-8130 canonical authenticator set (secp256k1, P-256, WebAuthn, delegate) is classical-only at Draft stage, with a non-secp256k1 self-authenticator slot reserved and a post-quantum verifier named only as an example occupant
- Gate 1a, Hybrid KEM: FAIL , no documented hybrid KEM at sequencer, batcher, proposer, TEE-prover or any operator-to-operator transport. Observed but not gate-qualifying: the issuer-operated public RPC edge mainnet.base.org, sepolia.base.org and mainnet.flashblocks.base.org negotiate TLS 1.3 with the hybrid X25519MLKEM768 group (live handshake 2026-08-19); this is CDN-edge termination, undocumented by Base, client-facing only, and the Alchemy Base endpoint negotiates classical X25519 only
- 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 Base team or corporate-parent position on what happens to dormant Base-native 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 the freeze automatically. Base-only-resident funds have no L1 fingerprint to freeze.
Seven dimensions
Each dimension scores 0–100 internally; the weighted roll-up produces the QRI.
1 Cryptographic Exposure weight 12% 31 / 100
Base inherits Ethereum's user-layer primitives and posts batches to Ethereum as EIP-4844 blob transactions signed by the batcher's ECDSA secp256k1 key. The issuer's smart-account product supports a second user-layer signing path, WebAuthn / passkey on secp256r1, verified by the P256VERIFY precompile that Base has had since the Fjord hardfork (RIP-7212, mainnet 2024-07-10 16:00 UTC); the Azul hardfork (mainnet 2026-05-28 18:00 UTC) only raised its gas cost from 3,450 to 6,900 to match EIP-7951 on L1. Azul replaced the interactive fault-proof game with a multi-proof system: permissioned TEE provers (AWS Nitro Enclaves) sign checkpoint output roots on the common path and permissionless ZK provers back or dispute them; every primitive in that system is classical. Still no canonical primitive-inventory document at base.org; the corporate parent committed on 2026-07-23 to a company-wide cryptographic inventory covering Base (announced, not yet published). The only post-quantum primitive observable anywhere near the stack is ML-KEM-768 inside the hybrid TLS key agreement at the CDN edge of the issuer-operated RPC; it is transport, not protocol, and Base has not documented it.
ECDSA secp256k1 (EOA signatures; batcher, proposer and operator L1 transactions) · ECDSA secp256r1 / P-256 (passkey / WebAuthn owners of the issuer's ERC-4337 smart accounts; verified by the P256VERIFY precompile at 0x100, present since Fjord via RIP-7212 and re-priced to 6,900 gas at Azul per EIP-7951; the issuer's WebAuthn library attempts the precompile first and falls back to the FreshCryptoLib Solidity verifier) · Keccak-256 (hashing / RLP / Merkle-Patricia trie) · SHA-256 (Ethereum consensus-layer SSZ hashing on the settlement chain; EVM precompile 0x02) · BLS12-381 (Ethereum L1 consensus signatures Base settles into; KZG commitments over BLS12-381 for the EIP-4844 blob batches Base posts since Ecotone) · ECDSA P-384 (AWS Nitro Enclave attestation certificate chain and COSE document signature for the Azul TEE provers; classical) · TLS 1.3 at the issuer-operated RPC edge: X25519MLKEM768 hybrid key agreement (X25519 + ML-KEM-768 per FIPS 203, IETF TLS hybrid draft group), AES-256-GCM or CHACHA20-POLY1305 record protection, ECDSA P-256 certificate signature (observed live 2026-08-19; CDN-terminated) ECDSA secp256k1→ Shor-break-via-DL-without-pairingsECDSA secp256r1 / P-256 (passkey)→ Shor-break-via-DL-without-pairingsECDSA P-384 (Nitro attestation chain)→ Shor-break-via-DL-without-pairingsKeccak-256→ Grover-weaken (256→128-bit preimage)SHA-256→ Grover-weakenAES-256-GCM / CHACHA20-POLY1305 (edge TLS)→ Grover-weaken (256→128-bit key search)X25519 component of X25519MLKEM768→ Shor-break-via-DLML-KEM-768 component of X25519MLKEM768→ PQ (FIPS 203, lattice); transport-only, CDN edgeBLS12-381 (Ethereum L1 consensus; EIP-4844 KZG)→ Shor-break-via-pairings (KZG blob commitments are ephemeral DA, ~18-day shelf life, not permanent state)
Base has zero post-quantum cryptography in production inside the protocol. No lattice, hash-based, code-based, or isogeny family has shipped at the account, execution, proof, batch or settlement layer. The ML-KEM-768 component observed in the hybrid TLS key agreement at the issuer's CDN edge is a transport default, not a protocol deployment, and earns no family-diversity credit.
No NIST FIPS 203/204/205 PQC primitive deployed by Base or its corporate parent inside the Base chain. (FIPS 203 ML-KEM-768 appears only in CDN-edge TLS, outside the protocol.)
OP Stack components Base inherited were audited by Trail of Bits, Sigma Prime, Spearbit and Cantina (public security-review index, reports 2020-2025). Base states every onchain component and proof system in Azul went through internal and external audits and ran an Immunefi audit competition (2026-04-21 to 2026-05-04, $250,000 maximum reward pool). The issuer's ERC-4337 smart-account contracts were audited by Cantina (2023-12, 2024-04), Certora (2024-02) and Code4rena (2024-03). secp256r1 verification runs through the protocol-level P256VERIFY precompile with a Solidity fallback. Tier 1 (classical ECC + Keccak-256 + SHA-256).
2 Quantum Recovery Exposure weight 8% 22 / 100
Base carries multi-billion-dollar value: TVL peaked at ~$5.58B on 2025-10-07 and stood at ~$4.7B on 2026-08-18 (public TVL aggregator, independently reconstructible). Both secp256k1 EOAs and passkey smart accounts are in active use; LayerQu has not measured the split. Passkey smart accounts (secp256r1) expose the P-256 public key on-chain from account creation (owner public keys are stored in the account) and inherit the same Shor-break exposure as EOAs that have spent.
Base has been open to the public since 2023-08-09 (36 months at evaluation date), old enough for material dormant-balance accumulation.
Batcher ECDSA secp256k1 L1 transactions, proposer checkpoint submissions, and TEE-prover checkpoint signatures are forgeable post-Shor against historical chain history; the permissionless ZK dispute path does not retroactively protect past settlements.
Issuer-operated public endpoints (mainnet.base.org, sepolia.base.org, mainnet.flashblocks.base.org) terminate TLS 1.3 at a CDN edge that negotiates the hybrid X25519MLKEM768 key agreement (X25519 + ML-KEM-768 per FIPS 203) with AES-256-GCM, observed live 2026-08-19; certificate authentication is ECDSA P-256 (classical). Alchemy's Base endpoint negotiated classical X25519 only and rejected a hybrid-only handshake; other third-party endpoints not tested. Flashblocks (200 ms preconfirmation stream, public mainnet launch 2025-07-16) is served over the same CDN edge. Sequencer, batcher, proposer and TEE-prover transports are undocumented; no operator-stated hybrid KEM policy exists. Scored above the baseline 5 because client-to-RPC transport for the operator's own endpoints is HNDL-mitigated in practice for clients that offer the hybrid group; capped well below mid-range because it is a CDN default, undocumented, and does not cover operator-internal transport.
3 Metadata, Anonymity & Confidentiality weight 8% 13 / 100
Pseudonymous transparent EVM ledger; full tx graph public. The issuer's passkey smart accounts create linkable user identities (passkeys are device-bound and onboarding commonly runs through the parent exchange's KYC'd app), which raises the de-facto identification rate for that account population.
Single operator-run sequencer (NC=1 for ordering). The operator's own postmortems record a 33-minute block-production halt on 2025-08-05 and halts of 116 and 20 minutes on 2026-06-25 and 2026-06-26 (sequencer block-building bug, stale journal state). Major public RPC endpoints include the issuer-operated mainnet.base.org, Alchemy and QuickNode. Transactions are submitted to the operator's sequencer; Flashblocks (since 2025-07) stream 200 ms preconfirmations publicly, increasing observability of pre-confirmation state.
Canonical Base Bridge (L1StandardBridge) deposits and withdrawals are direct on-chain links between Ethereum L1 and Base. Across, LayerZero, Hop and Stargate make L1↔L2 hops traceable. The parent exchange's direct-to-Base withdrawal pathway creates a CEX-internal linkage that, combined with KYC, is a strong de-anonymization vector.
No shielded pool on Base; transparent ledger. ECDSA Shor-break does not change tx-graph privacy (it was never private), but does enable retroactive recovery of any private key whose pubkey ever appeared on-chain.
No on-chain mixer integrated into the Base protocol.
4 Migration Architecture weight 15% 75 / 100
Base inherits Ethereum's hard-fork-driven crypto-agility. EIP-7702 is live on Base since the Isthmus hardfork (mainnet 2025-05-09 16:00 UTC), which ported the Pectra execution-layer EIPs two days after L1 Pectra (2025-05-07). The issuer's smart-account product already exercises algorithm flexibility at the user layer: a single ERC-4337 account can verify either secp256k1 or secp256r1 signatures. The 2026-02-18 move to a unified operator-run stack gave Base independent control of its hard-fork cadence (Base targets six hard forks per year), since exercised: Azul activated on mainnet 2026-05-28 and Beryl was scheduled for 2026-06-25. Native account abstraction per EIP-8130 (Draft, authored at the corporate parent) is previewing on the public Vibenet devnet and is announced for the Cobalt upgrade in September 2026; its authenticator-contract design admits new signature algorithms without account migration, but no post-quantum authenticator exists in the canonical set, and Base has not stated whether it adopts the Superchain's January 2036 ECDSA-EOA deprecation now that it has left the OP Stack.
The issuer's smart-account product is a concrete account-abstraction deployment: ERC-4337 contract wallet, passkey-based onboarding (no seed phrase), multi-owner architecture allowing both secp256k1 address owners and secp256r1 passkey owners on the same account. EIP-7702 EOA-delegation is live (Isthmus, 2025-05-09). Base has not adopted the Superchain 10-year EOA-to-PQ-AA roadmap (published 2026-01-14, targeting January 2036, subject to Superchain governance). Since 2026-05-22 the base/base monorepo carries an EIP-8130 implementation (native AA by account configuration; 68 commits and 8 open PRs as of 2026-08-19) whose account-state design reserves the self-authenticator slot for a non-secp256k1 verifier and gives a post-quantum verifier as the example occupant. It runs only on the public Vibenet devnet (chain ID 84538453, pre-Sepolia), Base announced it for the Cobalt upgrade in September 2026, and every authenticator in the current Draft's canonical set (secp256k1, P-256, WebAuthn, delegate) is classical.
Coordinated upgrade record over 36 months: public mainnet launch on the Bedrock OP Stack (2023-08-09); Canyon (2024-01-11), Delta (2024-02-22), Ecotone (2024-03-14), Fjord (2024-07-10), Granite (2024-09-11), Holocene (2025-01-09), Isthmus (2025-05-09), Jovian (2025-12-02) OP Stack hard forks; permissionless fault proofs (2024-10); Security Council control of upgrades (2025-04) and Stage 1 per Base's own statement; Flashblocks (mainnet 2025-07); unified-stack announcement (2026-02-18); Azul, described by Base as its first independent network upgrade, activated on mainnet 2026-05-28 (15 days after the 2026-05-13 target its announcement blog stated); Beryl scheduled for mainnet 2026-06-25 18:00 UTC with the required client release (v1.1.1) published on its Sepolia activation date, 2026-06-18 (LayerQu did not independently confirm Beryl's mainnet activation); Cobalt announced for September 2026.
The issuer's smart-account multi-owner architecture (secp256k1 address + secp256r1 passkey verifiable by the same account) is an existing template for parallel-signature verification at the smart-account layer. A hybrid (classical ECDSA + ML-DSA) signature path is architecturally constructible by adding a PQ verifier alongside the existing ones, and the EIP-8130 implementation makes that slot concrete in code (a reserved non-secp256k1 self-authenticator slot annotated with a post-quantum verifier as the example). That code is devnet-only, and the slot design is mutually exclusive with the inline secp256k1 self-authenticator, replacement rather than hybrid composition. No hybrid has been spec'd or shipped.
Base has no stateful hash scheme (XMSS/LMS/leanXMSS) at the protocol level; default-pass per v3.1.0 rule.
N/A. Base is a rollup with a single operator-run sequencer for transaction ordering. Since Azul, state validity rests on the AggregateVerifier multi-proof system on L1: permissioned TEE provers (AWS Nitro Enclaves) sign checkpoint output roots on the common path, permissionless ZK provers can back or dispute a proposal, and TEE-only or ZK-only proposals finalize after 7 days while TEE+ZK agreement finalizes in 1 day. There is no BFT consensus with BLS signature aggregation at the Base layer; consensus is inherited from Ethereum L1. Not scored, so it is left out of the dimension total rather than counted as a zero.
5 Deployment Execution weight 22% 11 / 100
Zero PQ signatures observed in Base's transaction stream. No Base contract or precompile shipped that exposes ML-DSA / ML-KEM / SLH-DSA / Falcon. The Azul hardfork (mainnet 2026-05-28) changed only classical primitives (EIP-7951 re-pricing of secp256r1). Transport-layer note, not scored here: the issuer-operated RPC edge negotiates X25519MLKEM768 hybrid TLS (CDN default, observed 2026-08-19); Alchemy's Base endpoint does not. This is not on-chain PQC traffic.
A live code search of the base/base monorepo (which carries the base-consensus consensus client and the base-reth-node execution client) on 2026-08-19 returns zero hits for ML-DSA, ML-KEM, Dilithium, Falcon, SPHINCS+, SLH-DSA and Kyber; the single hit for 'lattice' is a time-scheduling construct, not cryptography. No PQC code exists in any client Base runs.
The Base sequencer is operated by its corporate parent under standard secp256k1 keys. There is no validator set on Base proper; since Azul the proposer role is operator-run, TEE provers are permissioned and registered on-chain, and ZK provers are permissionless; all use Ethereum-format ECDSA keys.
VOIDED to 0 per v3.1 rule (5a = 0). No PQC-specific dated milestones published for the Base chain. The corporate parent's 2026-07-23 PQC roadmap carries corporate-level horizons (an automated PQ-capable signing pipeline within one year; full MPC for lattice signing schemes over the following two to three years) and commits to a cryptographic inventory covering Base, with a Base-specific plan stated as forthcoming; none are Base-chain milestones with dates. Base's own dated chain milestone in the window, the Cobalt upgrade announced for September 2026 carrying EIP-8130, names no post-quantum scheme.
Trailing-12-month PQ-adjacent announcements touching Base: 2. (1) The corporate parent's 2026-07-23 roadmap names Base in a company-wide cryptographic-inventory commitment and states Base inherits much of Ethereum's post-quantum protection through its L1 relationship; it names no primitive for Base, sets no Base date, and says a concrete Base plan is still to come. (2) Base's 2026-07-17 native-account-abstraction announcement lists 'Quantum Ready' among EIP-8130's features ('key rotation and multiple authentication schemes allow users to upgrade to post quantum authentication') while the canonical authenticator set is classical and no PQ authenticator exists; the label is forward-looking rather than a shipped property. Shipped PQ on Base mainnet: 0. The overstatement risks are the inheritance framing (what Base inherits today is a roadmap on the settlement layer, not a deployed primitive) and the 'Quantum Ready' label. Below the mechanical deduction threshold; one point lower than the single-announcement posture to reflect the unqualified label; monitored for the promised Base-specific plan and for Cobalt.
No PQ signature deployed; no mainnet bytes-per-block multiplier observable.
6 Supply Chain Vendor Readiness weight 25% 19 / 100
Top-3 (LayerQu selection): the issuer's own consumer wallet app (rebranded from the parent exchange's wallet app on 2025-07-16) together with its passkey smart-account product, MetaMask, Rabby. Hardware: Ledger, Trezor. Trezor Safe 7 ships post-quantum firmware integrity: ML-DSA-44 device-attestation certificate and SLH-DSA firmware-signature verification at the boot level (Trezor names 'SLH-DSA-128'; the firmware build compiles the SPHINCS+ reference code with the sphincs-sha2-128s parameter set, i.e. the SLH-DSA-SHA2-128s class), dual-signed with Ed25519; Trezor states transaction signing remains classical. No wallet vendor publishes a dated PQ roadmap at the smart-account verifier layer; passkey signing is secp256r1 (Shor-vulnerable).
Top-3 (LayerQu selection): canonical Base Bridge (L1StandardBridge), Across, LayerZero. Canonical Base Bridge inherits ECDSA secp256k1 + Keccak-256 from Ethereum L1 settlement and, since Azul, the AggregateVerifier TEE/ZK multi-proof path for withdrawals; no PQ verification path. Across settles relayer repayments through bonded bundle proposals secured by the UMA Optimistic Oracle on Ethereum L1, all ECDSA secp256k1; no PQ path. No public PQC roadmap located for LayerZero.
Top-3 (LayerQu selection): the parent company's custody arm (operator-affiliated and a major custodian of Base-native assets), BitGo, Fireblocks. On 2026-07-23 the corporate parent published its PQC roadmap: a post-quantum rebuild of the key-management system that protects approximately 99.9% of its custodied assets, an automated PQ-capable signing pipeline within one year (secure enclaves, secret-sharing, threshold cryptography, algorithm-agnostic), full MPC for lattice-based signing schemes over the following two to three years, and founding membership of the Bitcoin Security Consortium (nine founding members including BlackRock, Fidelity Digital Assets, Block, Blockstream and Strategy; $15M aggregate pledged over three years; supporting proposals such as BIP-360). This is a dated, named custodian PQC roadmap; it is announced, not shipped, and corporate-wide rather than Base-specific. BitGo published Bitcoin-wallet quantum-risk management tooling (2026-07-22: quantum risk score, exposed-address remediation, address-type defaults) and announced a post-quantum MPC transaction-signing simulation with Silence Laboratories (press release, 2026-05-26); neither is a dated PQC custody roadmap and neither covers EVM/Base assets. Fireblocks has published explanatory posts on the quantum threat (2025-07, 2026-04) but no dated PQC roadmap was located.
Top-3 RPC: issuer-operated mainnet.base.org, Alchemy, QuickNode. No published PQ TLS roadmap from any of the three; in practice the issuer-operated endpoints (mainnet.base.org, sepolia.base.org, mainnet.flashblocks.base.org) negotiate X25519MLKEM768 hybrid TLS 1.3 at their CDN edge (observed 2026-08-19), Alchemy's Base endpoint negotiates classical X25519 only, QuickNode endpoints are customer-specific and were not tested. HSM: AWS KMS has offered ML-DSA-44/65/87 signing keys (FIPS 204) since 2025-06; no Base-bound HSM PQ deployment is documented. TEE attestation chain: Base's Azul TEE provers run in AWS Nitro Enclaves; Nitro attestation documents are COSE-signed with ECDSA P-384 under an X.509 chain to the AWS Nitro root (classical), and Base verifies Nitro attestations in production via ZK proofs (operator engineering blog, 2026-08-11); no PQ attestation path is documented by the TEE vendor or by Base.
7 Governance & Coordination weight 10% 55 / 100
Sequencing centralized at its operator (NC=1 for ordering); batch submission, proposing and challenging are operator-controlled roles. Permissionless fault proofs launched 2024-10; Security Council control of contract upgrades from 2025-04; Base states it has reached Stage 1. Upgrades require a 2-of-2 of the operator's 3-of-6 signer multisig and an 8-of-11 Security Council multisig of 11 independent entities (roster as of 2026-02; an additional independent signer was added in lieu of the OP Stack maintainer at the 2026-02-18 stack change). Client diversity: Azul dropped support for all clients except base-consensus and base-reth-node, so Base now runs a single operator-maintained client pair.
36-month track record of coordinated upgrades through the operator's and the OP Stack maintainer's channels: Bedrock launch, nine OP Stack hard forks through Jovian (2025-12-02), permissionless fault proofs (2024-10), Security Council (2025-04), Flashblocks mainnet (2025-07), unified-stack announcement (2026-02-18), Azul mainnet activation (2026-05-28), Beryl scheduled (2026-06-25), Cobalt announced (2026-09), with a public client-release ledger in base/base (42 releases, v0.1.0 of 2025-04-28 through v1.2.0 of 2026-07-22) whose version bumps align with activation dates (v0.9.1 one day before Azul mainnet; v1.1.1 on Beryl's Sepolia date). The 2025-08-05 sequencer halt (33 minutes) and the 2026-06-25/26 halts (116 and 20 minutes) were each followed by a public postmortem. The move off the OP Stack was announced 2026-02-18 and completed at Azul without a contested fork.
A Nasdaq-listed corporate parent is the named protocol-development entity for Base. Jesse Pollak is the publicly identified lead for the Base chain/protocol; on 2026-07-15 he handed leadership of the separate Base App consumer product to another executive, a product-organization change with no reported change to chain-protocol governance. The corporate parent has a published mandate, conditional OCC approval to charter a national trust company (2026-04-02), and a corporate-grade incident-response apparatus. No designated PQ lead within the Base team; the 2026-07-23 corporate PQC roadmap is authored by the parent company's chief information security officer, a corporate role, not a Base-chain one.
The corporate parent is a publicly listed company with a mature security-incident response track record. Sequencer halts (2025-08-05; 2026-06-25/26) recovered without contested fork. No precedent of a coordinated cryptographic-primitive change under active attacker pressure on Base specifically.
No quantum canary (community honeypot, rate-limit, cryptographic tripwire, or automated response) embedded in Base protocol.
Source-disagreement disclosure
v3.1 requires every chain card to publish material divergences among authoritative sources, plus the delta-QRI under alternative weighting.
Until the 2026-02-18 announcement of a unified, operator-run stack, Base's path-to-PQ was effectively the OP Stack's path. Azul (mainnet 2026-05-28) completed the client consolidation (only base-consensus and base-reth-node are supported). Base now has independent control of its execution and proof system but has not yet articulated what it intends to do with that control regarding cryptographic primitives, and has not stated whether the Superchain's January 2036 ECDSA-EOA deprecation (published 2026-01-14, subject to Superchain governance) still applies to it.
The corporate parent's PQC posture is corporate, not chain-level: the 2026-07-23 roadmap details custody-system PQ engineering (a post-quantum rebuild of the key-management system, a one-year signing-pipeline target, a two-to-three-year lattice-MPC target) while committing Base only to a cryptographic inventory and inheritance of Ethereum's roadmap. None of it operates at the Base-chain level.
No primary source gives a dated quantum-proof institutional custody offering for late 2026. The verified 2026-07-23 roadmap gives within-one-year (automated PQ-capable signing pipeline) and two-to-three-year (full MPC for lattice signing schemes) horizons. This card carries only those horizons.
The operator's engineering blog (2026-04-21) announced a planned Azul mainnet activation of 2026-05-13; the canonical upgrade specification and the network configuration changelog record actual activation on 2026-05-28 18:00 UTC. The specification is treated as authoritative; the 15-day gap is an internal announced-vs-actual slip, not a third-party disagreement.
Base's network configuration changelog lists Flashblocks as enabled on mainnet on 2025-07-07; Base's engineering blog gives the public mainnet launch as 2025-07-16. Both are operator sources; this card cites 2025-07-16 as the public launch and notes the earlier enablement date.
The operator's initial status update implied a shorter halt; the operator's own postmortem on its status page states a 33-minute block-production halt on 2025-08-05. Some secondary coverage repeated earlier, shorter figures. This card uses the postmortem figure.
The Superchain post-quantum roadmap page carries a publication date of 2026-01-14; most press coverage, and the 2026-05-01 published reading, dated it 2026-01-26. This card uses the publisher's own date.
Base's 2026-07-17 native-account-abstraction announcement lists 'Quantum Ready' as a feature of EIP-8130 without naming a post-quantum scheme; the EIP's canonical authenticator set is classical. This card counts it as a PQ-adjacent announcement (announced_count 2, no deduction) rather than a primitive claim; a stricter reading would count it as zero announcements, a looser one as PQC-washing. Neither reading moves the score.
The hybrid X25519MLKEM768 key agreement observed at mainnet.base.org is a CDN-edge default that Base has not documented or claimed. This card credits it modestly in 2d (transport HNDL) and 6d, and not at all in 1c/1d/5a or the KEM gate, because it is neither a protocol primitive nor an operator-stated policy. A reviewer who treats CDN defaults as zero-credit would withhold the 2d and 6d credit entirely.
Delta-QRI under alternative weighting
Estimated −3 (QRI ≈ 24 if supply-chain weighted at 30%); estimated +2 (QRI ≈ 29) if the corporate custody roadmap were credited as a chain-level plan in Dim 5, which this card declines to do because no Base-chain primitive or date has been published; 0 if the CDN-edge hybrid TLS credit in 2d is removed.
Announcement-to-shipped ratio
Announced: 2. Shipped: 0. Ratio: 1.
Tag: announced-not-shipped, below deduction threshold. Two PQ-adjacent announcements touch Base in the trailing 12 months: the corporate parent's 2026-07-23 roadmap (names Base in a company-wide cryptographic-inventory commitment; names no Base primitive and sets no Base date) and Base's own 2026-07-17 native-account-abstraction announcement, which lists 'Quantum Ready' (key rotation and multiple authentication schemes 'allow users to upgrade to post quantum authentication') among EIP-8130's features while the canonical authenticator set contains no post-quantum scheme. Nothing post-quantum has shipped on Base mainnet. Neither announcement names a primitive, so neither is a press-release primitive claim in the strict sense of the 5e rule; with zero shipped the quotient is not numerically meaningful and is recorded at 1.0 with this flag rather than as an unbounded ratio. Re-check when the promised Base-specific plan publishes and when Cobalt activates.
Peers in the rollup-L2 profile
9 chains closest to Base by Stage then QRI.