TL;DR: Banks are migrating to NIST-approved post-quantum cryptography (PQC) to protect long-term financial data from “harvest-now, decrypt-later” quantum attacks. Start by inventorying crypto assets, then deploy hybrid schemes (classic + PQC) in TLS, code signing, and internal PKI using a phased rollout.
Step 1: Audit Your Cryptographic Inventory
Map every certificate, key exchange, and digital signature across core banking systems, ATMs, mobile apps, and interbank APIs. Use tools like openssl or commercial Crypto-Agility scanners to identify where RSA/ECC is used. Prioritize systems holding data with a shelf life >5 years (e.g., mortgage records, SWIFT messages).
If you want to dig deeper, check out our guide on Here are several SEO-optimized options, all under 70 charact.
Step 2: Select NIST-Approved PQC Algorithms
Use ML-KEM (FIPS 203) for key encapsulation (replacing ECDHE), ML-DSA (FIPS 204) for general signatures, and SLH-DSA (FIPS 205) for high-security, long-term attestation (e.g., firmware signing). For now, deploy in hybrid mode—e.g., X25519 + ML-KEM-768—to retain backward compatibility with legacy peers.
Step 3: Patch and Test in a Sandbox
Upgrade TLS 1.3 libraries (OpenSSL 3.5+, BoringSSL) and HSM firmware to support PQC. Run internal test transactions between two isolated nodes using hybrid cipher suites. Verify latency impact: ML-KEM adds ~0.1ms per handshake—negligible for banking, but test on your oldest hardware.
Step 4: Phased Production Rollout
Start with internal PKI and code-signing (lower risk), then move to customer-facing TLS endpoints after 2–4 weeks of stable monitoring. Use a feature flag to fall back to classical crypto if a vendor fails. Update your Certificate Authority to issue ML-DSA root and intermediate certs—but keep RSA roots alive until 2030 for old devices.
Step 5: Continuous Agility and Monitoring
Enable automated crypto-agility policies: scan weekly for any leftover classical-only sessions. Track NIST’s upcoming standards (e.g., FN-DSA) and be ready to swap algorithms without a full re-architecture. Train your security ops team on PQC key lifecycle—ML-KEM keys are shorter-lived than RSA.
Pro Tip: Never hardcode algorithm IDs. Use a crypto-config service that pushes PQC suites to all nodes simultaneously, and require dual signatures (ML-DSA + ECDSA) on high-value interbank settlement messages until 2028.
FAQ
Q: Will PQC break existing TLS connections with customers using old browsers?
A: No, if you deploy hybrid suites—the handshake negotiates classical crypto when the peer lacks PQC support. But you must keep RSA/ECDSA enabled during a 3–5 year transition window.
Q: What’s the biggest operational risk when migrating?
A: HSM firmware incompatibility—many legacy HSMs lack ML-KEM hardware acceleration. Verify your vendor’s PQC roadmap, and budget for HSM replacement before touching production payment rails.
Q: Can we use the same PQC certificate for internal and external systems?
A: Yes, but separate roots are safer: issue an internal SLH-DSA root for code signing and a public ML-DSA root for customer TLS, to limit blast radius if one is compromised.
Leave a Reply