The encryption underpinning nearly every secure connection on the internet has an expiration date, and for the first time, regulators have written it down. The National Institute of Standards and Technology finalized its first post-quantum cryptography standards in August 2024, major browsers have already begun negotiating quantum-resistant key exchange by default, and the US government has attached firm migration deadlines to federal systems. What was theoretical for two decades is now a procurement requirement and an operational deadline.
The math that quantum computers can break
Almost all public-key cryptography in use today rests on two problems that are hard for classical computers: factoring large integers, which underlies RSA, and computing discrete logarithms, which underlies Diffie-Hellman, ECDSA and EdDSA. Shor's algorithm, run on a sufficiently large fault-tolerant quantum computer, solves both efficiently. That would unravel key exchange in TLS, VPNs and SSH, along with the signatures behind code signing, public-key infrastructure and blockchain transactions.
Symmetric cryptography is far less exposed. Grover's algorithm only offers a quadratic speedup against ciphers like AES, so AES-256 and hash functions such as SHA-256 or stronger remain sound choices. The real migration burden sits entirely with public-key algorithms, which is why standards bodies have narrowed their focus there.
Why "harvest now, decrypt later" changes the timeline
No one can say precisely when a cryptographically relevant quantum computer will exist. But an adversary does not need one today to benefit from one tomorrow. Encrypted traffic intercepted now can simply be stored until decryption becomes feasible, a strategy known as harvest-now-decrypt-later. Anything protected only by classical key exchange today is already exposed to future compromise, which makes key exchange the first priority in any migration plan rather than an afterthought.
A useful way to judge urgency is to compare three numbers: how many years your data must remain confidential, how many years your migration will take, and how many years remain until a capable quantum computer arrives. If the first two exceed the third, the organization is already behind. US agencies including CISA and NSA have urged organizations to begin with a cryptographic inventory and vendor engagement now, rather than waiting for a definitive quantum milestone.
The new standards and what they replace
NIST's finalized standards give the industry concrete replacements. FIPS 203 defines ML-KEM, a lattice-based key encapsulation mechanism that replaces RSA and elliptic-curve Diffie-Hellman for establishing shared keys. FIPS 204 defines ML-DSA, the primary replacement for RSA and ECDSA digital signatures, also built on module lattices. FIPS 205 adds SLH-DSA as a backup signature scheme based on different mathematics entirely, in case weaknesses are later found in lattice-based approaches. A fourth standard, FN-DSA, is still in draft, and NIST selected a fifth algorithm, HQC, in March 2025 as a structurally distinct backup for key encapsulation.
These algorithms come with a practical cost: larger keys, larger ciphertexts and larger signatures than their classical predecessors. That size increase is already causing friction, from fragmented TLS handshakes to certificate fields that overflow and constrained devices that run short of memory. Most early deployments sidestep the risk of betting on one new algorithm by using hybrid key exchange, combining a classical method with a post-quantum one so the connection stays secure even if either approach is eventually broken.
Building a migration that does not need to be redone
NIST's draft guidance proposes deprecating 112-bit-security RSA and elliptic-curve algorithms after 2030, with a full disallowance of quantum-vulnerable schemes after 2035. Those dates remain in draft form, but US federal policy has already moved faster, with executive direction requiring agencies to migrate key establishment by the end of 2030 and digital signatures by the end of 2031, a timeline that extends to government contractors through procurement requirements.
- Assign clear ownership and add post-quantum readiness to vendor contracts and questionnaires.
- Build a cryptographic inventory covering protocols, libraries, certificates, hardware security modules and hard-coded algorithms.
- Prioritize externally reachable key exchange protecting long-lived data above short-lived internal signatures.
- Pilot hybrid key exchange on a subset of endpoints and test older proxies and load balancers against larger handshake messages.
- Separate algorithm choice from application code so future changes require configuration, not a rewrite.
Hardware vendors, particularly those supplying public-key infrastructure and security appliances, often have the longest lead times, so firmware roadmaps should be requested early rather than assumed.
A distinct problem for public ledgers
Blockchains face a harder version of the signature problem because their data is permanent. Any public key ever revealed on-chain remains visible indefinitely to a future attacker. In Bitcoin, older address formats and any reused address expose the underlying public key; a significant share of circulating supply already sits in such exposed outputs. Ethereum accounts reveal their public key the moment they send a transaction, placing most active accounts in a comparable position. Fixing this requires coordinated protocol upgrades and user migration, a process that historically unfolds over years, not months. Exchanges and custodians can act now by avoiding address reuse, tracking post-quantum signature proposals for the chains they support, and designing wallet infrastructure so signing algorithms can be replaced without rebuilding the system from scratch.