What is post-quantum cryptography and why it matters? Post-quantum cryptography (PQC) is a family of encryption and signature algorithms built to resist quantum computers while running on ordinary hardware. The risk is already live. Attackers can record your encrypted traffic today and decrypt it once a large quantum computer exists, a threat NIST calls harvest now, decrypt later. NIST finalized its first three PQC standards in August 2024, and its October 2026 transition report deprecates RSA and elliptic-curve cryptography after 2030. Browsers have moved first: over half of human web traffic through Cloudflare used post-quantum key exchange by October 2025.

What Is Post-Quantum Cryptography? The Short Definition

Post-quantum cryptography is the set of public-key algorithms, such as the ML-KEM and ML-DSA standards NIST published in August 2024, that stay secure against both classical and quantum computers.

You run post-quantum cryptography on the laptops, phones and servers you already own. It replaces the algorithms a quantum computer can break.

The Math That Quantum Computers Break

RSA depends on the difficulty of factoring large numbers. Elliptic-curve cryptography depends on the discrete logarithm problem. Peter Shor published a quantum algorithm in 1994 that solves both efficiently.

Moving from RSA to elliptic curves doesn’t help, because Shor’s algorithm breaks them equally. Diffie-Hellman key exchange falls to the same attack.

The replacements rest on other problems. ML-KEM and ML-DSA rely on Module Learning With Errors. The task is to recover a secret from noisy linear equations. FIPS 203 (2024) says ML-KEM is believed secure even against quantum adversaries. Believed is the word to watch, because post-quantum cryptography rests on an assumption nobody has broken yet.

Which Encryption Survives Untouched

AES-256 and SHA-3 stay outside the migration. Grover’s algorithm gives quantum computers only a quadratic speedup against symmetric ciphers. That leaves AES-256 with a 128-bit margin, according to Encryption Consulting’s 2026 summary of NIST’s transition schedule.

The job covers two things: key exchange and digital signatures. Bulk encryption keeps running as it does today.

So when a vendor calls a product quantum-safe, ask which of those two it replaced.

Why Post-Quantum Cryptography Matters Before Quantum Computers Exist

Post-quantum cryptography matters now because adversaries can store encrypted data today and decrypt it later, a threat NIST’s October 2026 transition report names as a main driver of urgency.

Harvest Now, Decrypt Later

The attack needs only storage. A government or criminal group copies encrypted traffic, parks it, and waits for a quantum computer to open it.

Your exposure depends on how long the data must stay secret. NIST’s October 2026 guidance says organizations might prioritize FIPS 203 first for data that is particularly sensitive or must stay secure for a long time.

Medical records stay sensitive for a patient’s lifetime, so healthcare data privacy teams carry more harvest risk than most. A session cookie that expires in an hour carries almost none.

How Fast the Qubit Estimates Are Falling

Drug discovery and finance get the headlines among the applications of quantum computing, yet code breaking sets the deadlines.

In 2019, Craig Gidney and Martin Ekerå estimated that factoring RSA-2048 would take 20 million noisy qubits and eight hours. Gidney’s May 2025 paper cut that to fewer than 1 million qubits and under a week.

Alice & Bob’s October 2026 estimate needs 22,593 physical cat qubits and 3.14 days, on a different qubit design. Those are designs on paper. Alice & Bob says its headline figure assumes no bit flips, and that one flip per hour raises the need to about 45,000 qubits.

The designs use different error correction, so the three numbers don’t sit on one trend line. Each new estimate still shortens your runway to adopt post-quantum cryptography, because migration is slow and the estimates only have to fall once more.

The Three NIST Standards and the 2030 and 2035 Deadlines

NIST finalized FIPS 203 (ML-KEM), FIPS 204 (ML-DSA) and FIPS 205 (SLH-DSA) on August 13, 2024, covering key establishment and digital signatures.

  • August 13, 2024: NIST publishes FIPS 203, 204 and 205.
  • November 12, 2024: NIST releases the initial draft of IR 8547, its transition timeline.
  • March 2025: NIST selects HQC, a code-based algorithm, as a backup key-establishment method (NIST IR 8545).
  • June 2026: Executive Order 14412 requires federal high-value assets to adopt post-quantum key establishment by 2030 and signatures by 2031.
  • October 27, 2026: NIST publishes the final IR 8547.
  • After 2030 and after 2035: quantum-vulnerable public-key algorithms become deprecated, then disallowed under NIST standards.

ML-KEM-768 is the default NIST recommends, with an encapsulation key of 1,184 bytes and a ciphertext of 1,088 bytes, per NIST’s FIPS 203 standard (2024). HQC gives the world a fallback built on different math, so one lattice break wouldn’t leave key exchange without a standard.

Deprecated means you can still use an algorithm but the risk is yours. Encryption Consulting’s October 2026 analysis puts it bluntly: deprecated doesn’t mean safe. Disallowed means a system conforming to NIST standards can’t use it.

Encryption Consulting’s analysis of the final report says the 2030 date now covers RSA-3072 and P-256, which the 2024 draft left alone until 2035. Verify that against the report text before you set an internal deadline.

NIST’s report also accommodates hybrid deployments that pair a classical algorithm with a post-quantum one, at the cost of more computation. Browsers already do this, which is how Cloudflare reached its over-half figure in October 2025.

Hybrid mode is the reason your 2026 connection can use post-quantum cryptography before any certificate changes.

Why Signatures and Certificates Are the Hard Half of the Migration

An ML-DSA-44 signature is 2,420 bytes, compared with 64 bytes for Ed25519, according to FIPS 204 (2024) and RFC 8032 (2017), which is why signatures slow migration more than key exchange.

Cloudflare’s August 2024 analysis of NIST’s standards calls signature migration less urgent than key agreement but much more challenging. Signatures matter from the day a capable machine exists. Recorded traffic matters from today.

The byte counts below come straight from the standards.

MeasureEd25519ML-DSA-44SLH-DSA-128s
Public key (bytes)321,31232
Signature (bytes)642,4207,856
Governing documentRFC 8032 (2017)FIPS 204 (2024)FIPS 205 (2024)
Broken by Shor’s algorithmYesNo known attackNo known attack
MeasureX25519ML-KEM-768
Key share sent (bytes)321,184
Reply received (bytes)321,088
Governing documentRFC 7748 (2016)FIPS 203 (2024)
Broken by Shor’s algorithmYesNo known attack

A browser checking a certificate chain verifies several signatures and receives several public keys, so every extra byte repeats on every new connection. SLH-DSA-128s keeps a 32-byte public key but produces 7,856-byte signatures, which makes it a poor fit for per-connection handshakes.

Encryption Consulting’s 2026 analysis says PKI, hardware security module and embedded-system replacement cycles routinely take years, which makes 2030 the date that constrains most plans. Devices that can’t be updated easily get the sharpest version of this problem. Take the sensors in smart city networks: a signature check built into firmware stays until the hardware is replaced.

If your product ships a 64-byte signature check in firmware today, that check is the part you can’t patch later.

Myths About Quantum-Safe Encryption That Waste Budget

Longer RSA keys don’t stop quantum attacks, because Peter Shor’s 1994 algorithm breaks RSA at any key length. The fix is a different algorithm family.

Post-quantum cryptography isn’t proven safe either. SIKE reached NIST’s fourth round. In August 2022, KU Leuven researchers Wouter Castryck and Thomas Decru broke it, recovering keys in about 62 minutes on one classical CPU core, as The Hacker News reported. Rainbow had fallen in February 2022.

Neither algorithm was standardized, and the attack didn’t touch the lattice schemes. FIPS 203 still says NIST will reevaluate ML-KEM every five years, which is the practical argument for hybrid deployments.

Nothing is due before a quantum computer exists is the third misreading. Executive Order 14412 sets 2030 for key establishment on federal high-value assets. NIST’s October 2026 news release says organizations that work with the government will feel it too.

Who Should Migrate First, and a 30-Day Starting Plan

Teams that hold data with a long shelf life, such as medical records or legal archives, should move key exchange to ML-KEM first, while firmware makers should start with signatures.

Three situations change the order of your post-quantum cryptography work.

Long-lived secrets come first. Cloudflare’s March 2025 report describes its protection as TLS 1.3 with hybrid ML-KEM key exchange. A service still on TLS 1.2 probably needs that upgrade first.

Firmware comes second. A verification key written into a device in 2026 may still be checking updates in 2036. Firmware makers should weigh hash-based SLH-DSA from FIPS 205 against ML-DSA now.

Short-lived traffic can wait. Session data that loses value within hours carries little harvest risk, though the certificates protecting it still sit on the 2030 clock.

Teams already rebuilding access around a zero trust architecture have a head start, because that work already forces a list of every connection and certificate.

The scale is real. OMB’s July 2024 report put the cost of migrating non-national-security federal systems at $7.1 billion, as cited in a 2025 Computing Research Association paper. Thirty days is enough to start your post-quantum cryptography inventory.

  1. Inventory. List every use of RSA, ECDSA, ECDH and Diffie-Hellman, including certificates, VPNs, SSH keys and code signing.
  2. Rank by shelf life. Mark each item by how many years its data must stay secret.
  3. Test your browser. Open pq.cloudflareresearch.com, the check page Cloudflare published in August 2024, to see whether your browser negotiates post-quantum key agreement.
  4. Upgrade TLS. Move endpoints from TLS 1.2 to TLS 1.3.
  5. Ask vendors for dates. Request each supplier’s dated plan for FIPS 203 and 204 support.

Which of your systems holds a key you can’t rotate?

People Also Ask

What is post-quantum cryptography in simple terms?

Post-quantum cryptography is a new set of encryption and signature methods designed to stay secure against quantum computers. The methods run as software on ordinary computers. They replace RSA and elliptic-curve algorithms, which Shor’s algorithm breaks. NIST published the first three standards, FIPS 203, 204 and 205, on August 13, 2024.

Does quantum computing break all encryption?

Quantum computing does not break all encryption: it breaks public-key algorithms such as RSA and elliptic curves, while AES-256 survives. Shor’s algorithm does the breaking, and Grover’s algorithm only gives a quadratic speedup against symmetric ciphers, per Encryption Consulting’s 2026 summary of NIST’s schedule. Migration targets key exchange and signatures.

What is harvest now, decrypt later?

Harvest now, decrypt later is an attack where someone records encrypted data today and decrypts it after a quantum computer becomes available. NIST’s October 2026 report names it as a main reason the transition is urgent. Data that must stay secret for many years, such as medical or government records, carries the most risk.

When could quantum computers break current encryption?

Nobody can name a date when quantum computers will break RSA-2048, but published resource estimates keep falling. Gidney’s May 2025 paper puts RSA-2048 under 1 million noisy qubits and under a week, and Alice & Bob’s October 2026 estimate cites 22,593 physical cat qubits and 3.14 days. No one has built machines that size yet, per PostQuantum.com’s October 2026 report.

What is the difference between post-quantum cryptography and quantum cryptography?

Post-quantum cryptography is software based on new math that runs on classical computers, while quantum cryptography, such as quantum key distribution, uses quantum physics and specialized hardware. PQC replaces RSA and elliptic-curve algorithms in existing systems. QKD exchanges keys over dedicated optical links. Encryption Consulting’s 2026 primer separates the two this way.

Frequently Asked Questions

Which post-quantum algorithms has NIST standardized?

NIST has standardized three algorithms and picked a fourth for a future standard. FIPS 203 specifies ML-KEM for key establishment, FIPS 204 specifies ML-DSA for signatures, and FIPS 205 specifies SLH-DSA, a hash-based signature scheme, all published August 13, 2024. In March 2025, NIST selected HQC as a backup key-establishment method built on error-correcting codes. ML-KEM-768 is NIST’s recommended default parameter set, per FIPS 203. Expect ML-DSA to appear in certificates first, and SLH-DSA where you want a signature that doesn’t depend on lattices.

Can I use post-quantum encryption today?

Yes, for key exchange, and you may already be using it. Cloudflare reported in October 2025 that over half of human-initiated traffic to its network used post-quantum key exchange, and PostQuantum.com notes that an up-to-date browser probably negotiates it already. You can check yours at pq.cloudflareresearch.com, a test page Cloudflare published in August 2024. Signatures lag, because Cloudflare’s August 2024 post expected signature adoption to take longer. For your own servers, TLS 1.3 support is usually the first gate, since the hybrid ML-KEM deployment Cloudflare describes runs on TLS 1.3.

Does post-quantum cryptography slow down websites?

Key exchange adds bytes more than computation. ML-KEM-768 sends a 1,184-byte encapsulation key and receives a 1,088-byte ciphertext, compared with 32 bytes each way for X25519, per FIPS 203 (2024) and RFC 7748 (2016). NIST’s August 2024 announcement credits FIPS 203 with comparatively small keys and fast operation among post-quantum options. Signatures hurt more: an ML-DSA-44 signature is 2,420 bytes against 64 for Ed25519, so certificate chains grow. The effect depends on your traffic mix, because a site serving many short connections notices extra bytes sooner than one holding long sessions.

Do small businesses need to worry about post-quantum cryptography?

Small businesses inherit most of the work through vendors, so your main task is asking them questions. Your cloud host, VPN, email provider and payment processor control the algorithms. Ask each one for its dated plan to support FIPS 203 and 204 and to retire RSA and ECDSA. NIST’s October 2026 report says organizations that work with the federal government will be affected as well. If you handle records that must stay private for many years, raise the question sooner. If you don’t, a yearly check-in with your vendors is probably enough.

What does deprecated after 2030 actually mean?

Under NIST IR 8547, deprecated means an algorithm such as RSA or ECDSA is still usable but carries risk the user accepts, and disallowed means NIST standards no longer permit it. The final report, published October 27, 2026, sets deprecation after 2030 and disallowance after 2035. For federal high-value systems, Executive Order 14412 adds harder dates: 2030 for key establishment and 2031 for signatures. If you sell to government agencies, those dates reach you through procurement language even when you aren’t a federal system yourself.

Ahmed UA

A technology journalist with over 13 years of industry experience covering AI, cybersecurity, mobile technology, gadgets, and global tech trends. He founded iCONIFERz in 2019 as a platform dedicated to making technology accessible to everyone — without the jargon. Follow Website, Facebook & LinkedIn.

Stay in the loop

Subscribe to our free newsletter.

You can unsubscribe anytime.

  • Integrating blockchain into legacy systems is usually an integration project, not a replacement project. Most organizations keep their ERP, CRM, databases, and business applications while adding blockchain where shared trust, auditability, and multi-party verification matter. The difficult part is rarely the blockchain itself. Data mapping, reconciliation, governance, security controls, and workflow changes consume far more time than smart contract development in most enterprise deployments. Why Companies Are Adding Blockchain to Existing Infrastructure Most enterprises have invested millions in software that [...]

KEEP READING

Latest Post