QianHeng乾珩 PQC Docs Hub量子文档 ✦ Ask AI✦ 问问文档 ⚐ Scan⚐ 扫一扫

Finance & Banking PQC Solution

Financial systems combine long-retention secrets, regulated cryptography, and a deep dependence on Shor-vulnerable public-key algorithms. That makes banking one of the first sectors that must act on post-quantum migration — both to defeat "harvest now, decrypt later" attacks and to stay compliant with China's commercial-cryptography regime.

Why act now

The drivers in finance are simultaneously a threat and a compliance obligation:

  • HNDL on long-lived data. Account records, transaction histories, KYC/identity material, and contracts retain value for 10–30 years. An adversary can record today's TLS/IPsec sessions or exfiltrate ciphertext now and decrypt it once a cryptographically relevant quantum computer (CRQC) exists. By Mosca's inequality (X+Y>Z), if secrecy lifetime plus migration time exceeds the time to a CRQC, you are already exposed.
  • 商用密码合规 & 密评. Under the Cryptography Law (effective 2020-01-01, SCA), commercial-cryptography application security assessment (密评, per GB/T 39786-2021 and GM/T 0115/0116-2021) is mandatory for CII, 等保 level-3+ systems, and important government/finance systems, with annual re-assessment.
  • CII crypto requirements. GM/T 0133-2024 (effective 2025-07-01) raises crypto requirements for critical information infrastructure — most large banks operate CII.
  • Sector & cross-border rules. PBoC / financial-industry JR/T standards govern domestic crypto usage; PCI DSS and GDPR apply to cross-border payment and data flows. 等保三级 (MLPS level 3) applies to core financial systems.

Risk & asset map

Symmetric primitives (SM3/SM4/ZUC, AES) are only weakened by Grover and are mitigated with larger sizes (AES-256, ZUC-256). The urgent targets are the Shor-broken asymmetric layers (SM2/SM9, RSA/ECC/ECDSA).

Asset / systemCrypto in useQuantum riskPriority
Payment & transaction systemsTLS, SM2/ECDSA signatures, SM2/ECDH key exchangeShor-vulnerable asymmetric layer; HNDL on session keysHighest
Core banking ledgerSM2/RSA signing, TLS to subsystemsShor-vulnerable signing & transportHigh
Financial HSM / key managementSM2/RSA key wrapping, ECDSAShor-vulnerable asymmetric wrapping; firmware-boundHigh
CBDC / digital-currency walletsSM2 / ECDSA signatures, SM9 IBCShor-vulnerable — long-lived wallet keysHigh
Cross-border payment & interbankTLS, RSA/ECDSA PKI, message signingShor-vulnerable; HNDL on intercepted trafficMedium
Long-retention records & archivesStored ciphertext (SM4/AES), SM2-wrapped keysGrover-only data, but Shor-vulnerable key-wrap — classic HNDLMedium–High
Symmetric bulk encryptionSM4 / AES-128 / ZUCGrover-only — move to AES-256 / ZUC-256Low (size bump)

Recommended architecture

The design goal is crypto-agility: replace hard-coded algorithms with negotiable, swappable suites so future parameter changes do not require re-architecting. Build on the patterns in construction-blueprint.html.

  • Hybrid key exchange. Combine a classical exchange (ECDH or SM2) with ML-KEM via a KDF, so a session stays secure if either component holds. This preserves 国密 compliance while adding quantum resistance — see hybrid.html.
  • Dual signatures. Sign with both a classical scheme (SM2 or ECDSA) and ML-DSA, giving verifiable backward compatibility and post-quantum integrity.
  • PQC-ify PKI & HSM. Issue hybrid/PQC certificates, add PQC support in HSM firmware and key-management, and plan for larger key/signature sizes. See pki.html.
  • High-assurance reference. CNSA 2.0 (ML-KEM-1024 + ML-DSA-87) is a useful upper-bound profile for the most sensitive flows.
  • Inventory first. No migration without a cryptographic bill of materials — see discovery.html and the quantum-risk-model.html.

Phased roadmap

PhaseWindowFocusKey outcomes
Phase 1Now – 18 moDiscovery, 密评 readiness, pilotCryptographic inventory & risk map; 密评 gap analysis against GB/T 39786; pilot hybrid key exchange on a payment/transaction channel; crypto-agility framework.
Phase 218 – 36 moCore banking, HSM & PKIDual signatures in core ledger; PQC-capable HSM firmware; hybrid/PQC PKI issuance; CBDC/wallet key strategy.
Phase 336 mo+Peripheral, cross-border, decommissionExtend to cross-border & interbank links; AES-256/ZUC-256 symmetric uplift; decommission legacy RSA/ECC and pure-classical certificates.

Reference cases

Synthesized from the 2025 安全牛 report on the cn-vendors.html page: 银河证券 ran an anti-quantum pilot (with Koal / 格尔软件), and a digital-currency exchange deployed a hybrid scheme combining classical and PQC key exchange. These confirm that hybrid-first, pilot-led migration is the practical CN-finance path.

KPIs & success criteria

  • 100% cryptographic asset coverage in the inventory; every Shor-vulnerable asset risk-scored.
  • 密评 readiness with no critical findings against GB/T 39786 / GM/T 0133-2024.
  • Hybrid key exchange live on all external payment/transaction channels by end of Phase 2.
  • Dual-signature coverage on core-banking and CBDC flows; PQC-capable HSM/PKI in production.
  • Legacy pure-classical asymmetric crypto retired on a published decommission schedule.

Standards & references

Warning
HNDL means the clock has already started for long-retention financial data. Any payment, transaction, or archived ciphertext protected today only by SM2/RSA/ECC key exchange should be treated as recordable-now, decryptable-later — prioritize hybrid key exchange on those channels first.

金融与银行后量子解决方案

金融系统集长期保密数据、受监管的密码以及对易受 Shor 攻击的公钥算法的深度依赖于一身,这使银行业成为最先必须行动的后量子迁移行业之一——既要抵御先收集、后解密(Harvest Now, Decrypt Later, HNDL)攻击,也要满足国家商用密码合规要求。

为何现在就要行动

金融行业的驱动力既是威胁,也是合规义务:

  • 长效数据面临 HNDL。账户记录、交易历史、KYC 身份资料以及合同的价值可保持 10 至 30 年;攻击者今天即可记录 TLS/IPsec 会话或窃取密文,待具备密码学意义的量子计算机(CRQC)出现后再行解密。根据 Mosca 不等式(X+Y>Z),若保密寿命加迁移时间超过 CRQC 出现的时间,你就已经处于暴露状态。
  • 商用密码合规与密评。依据《密码法》(2020-01-01 起施行,主管机构为 SCA),商用密码应用安全性评估(密评,依据 GB/T 39786-2021GM/T 0115/0116-2021)对关键信息基础设施(CII)、等保三级及以上系统以及重要政务与金融系统属强制要求,且需每年复评。
  • CII 密码要求。GM/T 0133-2024(2025-07-01 起施行)提升了关键信息基础设施的密码要求,而多数大型银行均属 CII。
  • 行业与跨境规则。人民银行与金融行业 JR/T 标准规范境内密码使用,跨境支付与数据流动还需满足 PCI DSSGDPR;核心金融系统适用等保三级。

风险与资产图谱

对称原语(SM3/SM4/ZUC、AES)仅被 Grover 削弱,可通过加大长度(AES-256、ZUC-256)缓解;紧迫目标是会被 Shor 攻破的非对称层(SM2/SM9,RSA/ECC/ECDSA)。

资产 / 系统所用密码量子风险优先级
支付与交易系统TLS、SM2/ECDSA 签名、SM2/ECDH 密钥交换易受 Shor 攻击的非对称层;会话密钥面临 HNDL最高
核心银行账务SM2/RSA 签名、子系统间 TLS易受 Shor 攻击的签名与传输
金融 HSM 与密钥管理SM2/RSA 密钥包装、ECDSA易受 Shor 攻击的非对称包装,且与固件绑定
CBDC 与数字货币钱包SM2/ECDSA 签名、SM9 标识密码易受 Shor 攻击,钱包密钥长期有效
跨境支付与行间交互TLS、RSA/ECDSA PKI、报文签名易受 Shor 攻击,截获流量面临 HNDL
长期留存记录与归档存储密文(SM4/AES)、SM2 包装密钥数据仅 Grover 相关,但密钥包装易受 Shor 攻击,典型 HNDL中至高
对称批量加密SM4 / AES-128 / ZUC仅 Grover 相关,迁移至 AES-256 / ZUC-256低(加大长度)

推荐架构

设计目标是密码敏捷性:以可协商、可替换的算法套件取代硬编码算法,使未来调整参数无需重构架构。可基于 construction-blueprint.html 中的模式构建。

  • 混合密钥交换。通过 KDF 将经典交换(ECDHSM2)与 ML-KEM 结合,只要任一组件未被攻破,会话即安全;这在增加抗量子能力的同时保留国密合规,详见 hybrid.html
  • 双重签名。同时使用经典方案(SM2ECDSA)与 ML-DSA 签名,兼顾可验证的向后兼容与后量子完整性。
  • PKI 与 HSM 后量子化。签发混合或纯 PQC 证书,在 HSM 固件与密钥管理中加入 PQC 支持,并为更大的密钥与签名尺寸预留余量,详见 pki.html
  • 高保障参考。CNSA 2.0(ML-KEM-1024 + ML-DSA-87)可作为最敏感流程的上限参考画像。
  • 先盘点。没有密码物料清单就不要迁移,参见 discovery.htmlquantum-risk-model.html

分阶段路线图

阶段时间窗重点关键产出
第一阶段当下至 18 个月盘点、密评就绪、试点密码资产盘点与风险图谱;对照 GB/T 39786 的密评差距分析;在某条支付交易通道试点混合密钥交换;搭建密码敏捷框架。
第二阶段18 至 36 个月核心银行、HSM 与 PKI核心账务引入双重签名;HSM 固件支持 PQC;签发混合或 PQC PKI;制定 CBDC 与钱包密钥策略。
第三阶段36 个月以上外围、跨境、退役扩展至跨境与行间链路;对称层提升至 AES-256/ZUC-256;退役遗留 RSA/ECC 与纯经典证书。

参考案例

综合 2025 年安全牛报告(见 cn-vendors.html 页面):银河证券联合格尔软件(Koal)开展了抗量子试点,某数字货币交易所部署了结合经典与 PQC 密钥交换的混合方案。这些案例印证了"混合优先、试点引领"是国内金融迁移的务实路径。

KPI 与成功标准

  • 盘点覆盖率达 100%,每项易受 Shor 攻击的资产均完成风险评分。
  • 对照 GB/T 39786 / GM/T 0133-2024 达到密评就绪,无重大问题。
  • 第二阶段末,所有对外支付交易通道均启用混合密钥交换。
  • 核心银行与 CBDC 流程实现双重签名覆盖,HSM 与 PKI 支持 PQC 并投产。
  • 按公布的退役时间表淘汰遗留的纯经典非对称密码。

标准与参考

警告
HNDL 意味着长期留存的金融数据,其倒计时已经开始。任何今天仅由 SM2/RSA/ECC 密钥交换保护的支付、交易或归档密文,都应视为"今天可被记录、未来可被解密",应优先在这些通道率先部署混合密钥交换。
⚑ Report an error⚑ 纠错与校正