Prof. Dr. Dennis-Kenji Kipker is a cybersecurity expert and works as Scientific Director of the cyberintelligence.institute.getty​Quantum computers will eventually make today’s public-key cryptography breakable. That’s not hypothetical for the financial sector; it’s an operational risk across payment systems, core banking infrastructure and long-lived customer records. The technical countermeasure, post-quantum cryptography (PQC), already exists. The real challenge facing bank and insurer CIOs and CISOs is migrating decades-old, interdependent technology stacks before the window closes.Modern cryptography rests on two mathematical problems that are computationally difficult for classical computers at sufficiently large sizes: factoring large numbers and computing discrete logarithms. Rivest-Shamir-Adleman (RSA) and elliptic-curve cryptography (ECC), widely used for key establishment and digital signatures in systems such as Transport Layer Security (TLS), depend on the continued difficulty of these problems.A powerful enough quantum computer running Shor’s algorithm would solve both, unraveling RSA and ECC outright. PQC replaces those foundations with different math—lattice problems, hash-based signatures and code-based schemes—that remain hard even for quantum computers.Crucially, PQC runs entirely on classical hardware and slots into existing protocols like TLS and IPsec, without an institution needing to own quantum technology. In my experience, that’s the detail executives find most reassuring: This is a software and certificate migration, not a hardware build-out.The urgency is sharpest for data that must remain confidential for years or decades, such as loan agreements, customer files and regulatory archives, so data intercepted today is already at risk, even if a capable quantum computer doesn’t arrive for another decade. This is a pattern known as “harvest now, decrypt later.”PQC is also frequently confused with quantum key distribution (QKD). QKD exchanges keys via the physical properties of photons over dedicated fiber links and works only point-to-point over limited distances, which is not commercially practical for core banking and payment rails.The standards are set.In August 2024, NIST finalized the first three PQC standards: FIPS 203, FIPS 204 and FIPS 205. ML-KEM handles quantum-safe key exchange; ML-DSA and SLH-DSA cover digital signatures. NIST plans to retire the weakest quantum-vulnerable algorithms after 2030 and phase out most others by 2035, and has a new algorithm, HQC, ready as a hedge against weaknesses in lattice-based schemes.No algorithm stays trusted forever, so cryptographic agility matters as much as any single standard. Systems that hard-code encryption algorithms are hardest to migrate; those calling cryptography through centrally managed libraries can swap algorithms as new findings demand it. Having walked several institutions through this kind of assessment, I’ve found this the most underestimated part of the project—legacy applications often bury cryptographic parameters deep in source code, turning a configuration change into a rewrite.For the multiyear transition, hybrid schemes—pairing a classical algorithm like RSA or ECC with a post-quantum one in the same handshake—are the practical answer: The connection stays secure as long as either algorithm holds. TLS 1.3 already supports this in production; Google enabled it by default in Chrome in 2020, and major cloud providers did the same. Institutions don’t need full market maturity before piloting hybrid deployments on their most exposed interfaces.Regulators are setting the clock.Washington has moved this from best practice toward expectation. The NSA’s Commercial National Security Algorithm Suite 2.0 calls for national security systems to transition away from RSA, Diffie-Hellman and elliptic-curve cryptography as CNSA 2.0 becomes mandatory in the coming years.CISA, the NSA and NIST jointly urge organizations supporting critical infrastructure, including financial services, to build a cryptographic inventory and quantum-readiness road map now, warning that adversaries may already be harvesting data for later decryption. FS-ISAC, representing financial institutions and infrastructure globally, has separately warned about “crypto-procrastination,” arguing that underestimating the complexity of migration or treating quantum as a distant risk could leave organizations scrambling to complete a transition that will take years.Build a quantum-safe architecture.Cryptographic agility only works if key and certificate management is centralized and decoupled from individual applications. In my experience, that decoupling is usually missing because systems built up over many years implemented cryptography inconsistently, application by application.A resilient public-key infrastructure with clear ownership and automated certificate rotation is the foundation every quantum-safe architecture needs—reinforced by zero-trust principles, which authenticate every session individually rather than trust a network perimeter. Long-lived credentials and signing keys are prime harvest-now, decrypt-later targets, so the two approaches strengthen each other.New systems should be built quantum-secure-by-design, and procurement should require vendors to commit to concrete PQC road maps before contracts are signed. A core banking system’s transaction signing often runs through a hardware security module whose manufacturer has no announced post-quantum timeline—a dependency that needs to surface early. The same applies to payment interfaces on ISO 20022, open banking APIs and cloud TLS endpoints—the most exposed and hardest to inventory—so institutions should push vendors to offer hybrid TLS and quantum-safe key management as a service.One of the biggest obstacles is organizational, not technical. I have repeatedly seen security architecture, application development and business units start with different priorities, and shared urgency takes real work to build before migration can start. A cross-functional program with clear governance—budget, timeline and ownership visible across several years—consistently outperforms treating PQC as a job for a single IT team. In-house cryptography expertise matters, too, since practitioners with real post-quantum experience remain scarce.Where can you start?Every migration should begin with a complete cryptographic inventory: algorithms, key lengths, certificate lifetimes, APIs and vendor dependencies. From there, prioritize by two factors: the sensitivity and retention period of the data, and how exposed a system is externally. Internet-facing interfaces holding long-lived customer data should migrate ahead of internal back-office systems, while embedded components like hardware security modules—dependent on vendor cycles outside an institution’s control—need the longest lead time.Post-quantum cryptography isn’t a project with an end date. It touches architecture, governance, procurement and workforce planning at once, running across multiple budget cycles, and succeeds where the board, CIO, CISO and chief risk officer own it jointly rather than treating it as a side initiative.Institutions that start the inventory now, treat crypto-agility as an architecture principle and pilot hybrid deployments today will move through the quantum transition on their own schedule, not someone else’s.Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?