Post-quantum signatures and standards: where NIST, ISO and the BSI stand at the end of 2026
Two signature standards finished, one in progress, nine candidates in the NIST round and a new ISO standard: the status for decision-makers.

- Since August 2024 there have been two final NIST standards for signatures; a third with smaller signatures is still in progress.
- In May 2026 NIST selected nine candidates for a third round of additional signature schemes; it will take about two years.
- In June 2026 ISO standardised three post-quantum key agreement schemes; the BSI continues to recommend hybrid use.
Contracts, invoices, software updates, certificates and approvals in the data room: wherever authenticity matters, digital signatures are involved. Unlike encryption, post-quantum signatures are not settled by a single standard. Standardisation continues on several levels, and anyone planning long-lived systems should know where things stand.
Why signatures behave differently from encryption
With encryption, the attack works retroactively: whoever records encrypted data today can decrypt it later with a powerful quantum computer. Signatures are different. A signature that was verified as valid today does not become invalid afterwards. It becomes dangerous as soon as an attacker with such a computer can forge new signatures, for example on software updates or certificates.
Two things follow: migrating signatures is less urgent than migrating key agreement. But it takes longer, because signatures are deeply embedded in infrastructure — in certificate chains, hardware security modules, smart cards, firmware and archives with retention periods. And where signatures must retain evidential value for many years, as with contracts, every year of lead time counts.
What is finished at NIST and what is not
In August 2024 the US NIST published two final signature standards: FIPS 204, based on lattice problems, and FIPS 205, based on hash functions (NIST). They deliberately complement each other: the first is fast and compact, the second rests on particularly well-understood assumptions but produces much larger signatures.
A third standard with particularly small signatures has been announced but is not yet final. It is considered demanding to implement securely. Until it is available, it belongs in the lab, not in production.
The third round of additional signature schemes
In parallel, NIST is looking for further signature schemes. Two goals stand out: smaller signatures for applications with little space, such as smart cards or embedded systems, and more diversity in the mathematical foundations. On 14 May 2026, after 18 months of evaluation, NIST selected nine candidates for the third round (announcement in the NIST PQC forum). They come from four families: multivariate schemes, "MPC in the head" schemes, a lattice scheme and an isogeny-based scheme. The round is expected to last about two years; the next NIST standardisation conference is planned for 2027.
For companies this means: new standards from this round are not expected before the end of the decade. Whoever migrates today uses the finished standards and plans so that further schemes can be added later.
Rule of thumb: Signature standardisation is not an end point but a process. Plan today with the finished standards and keep your systems open for the next ones.
What ISO has contributed
Alongside NIST, the international organisation ISO, together with IEC, also standardises post-quantum cryptography. In June 2026 the amendment ISO/IEC 18033-2:2006/Amd 2:2026 was published. It adds three key agreement schemes: the scheme standardised by NIST in FIPS 203 and two particularly conservative schemes that European authorities such as the BSI have recommended for years for applications with high protection needs (background).
That standards are not the last word is shown by news from this year: due to advances in cryptanalysis, the BSI now advises against using one of the two conservative schemes in new developments for the time being. With the recommended parameters a practical attack is not possible, and in hybrid use at least classical security is preserved (it-daily). For users it is an object lesson in crypto agility: algorithms must be replaceable without rebuilding the application. More in the article Crypto agility.
What the BSI recommends
The BSI publishes its recommendations annually in Technical Guideline TR-02102. For post-quantum algorithms the rule throughout is: use them hybrid, i.e. together with a proven classical algorithm. For classical asymmetric algorithms on their own, the BSI has set an end date: end of 2031, and end of 2030 for applications with particularly high protection needs (BSI TR-02102).
What you can do now
- Inventory your signatures: where do you sign today (code, documents, certificates, e-mails, devices)? How long must these signatures remain valid?
- Long-lived first: systems with long lifetimes (firmware, smart cards, archives) need the most lead time.
- Ask vendors: do libraries, hardware security modules and signing services support FIPS 204 and 205? When?
- Plan hybrid: classical and post-quantum signatures in parallel until all recipients have migrated.
- Ensure replaceability: treat algorithms as configuration, not as fixed code.
PQVault and signatures
PQVault uses NIST-standardised post-quantum algorithms in all modules and is built to be crypto-agile. Signatures and approvals in PQVault Spaces and sender verification in PQVault Seal are based on them. For your own applications the PQVault SDK offers the same building blocks. An overview of the basics is given in the article NIST standards for post-quantum cryptography.
PQVault is a product of QuantiCor Security GmbH, Darmstadt — specialists in post-quantum cryptography. More about QuantiCor Security


