SecurityRisk & complianceAnalysis
What NIST IR 8547, final on 27 October, and the US, NSA, UK, EU and Indian timetables say about retiring RSA and elliptic-curve cryptography, as of 2 November 2026.
Almost every encrypted connection and digital signature in use today rests on RSA or elliptic-curve cryptography, which a large enough quantum computer could break. Until last week the dates for retiring them were a draft. They are now final for the US government, and the UK, the EU and India have published timetables of their own. The question for a buyer is which of these dates touches the products you are buying now, and what to ask the vendor before you sign.
NIST published the final version of IR 8547, Transition to Post-Quantum Cryptography Standards, on 27 October 2026. Its central sentence reads: “NIST will deprecate quantum-vulnerable public-key cryptography after 2030 and disallow its use under NIST standards and guidelines after 2035.”12 In NIST’s terms, deprecated means the algorithm “may be used, but there is some security risk”, and the data owner decides whether to accept it. Disallowed means it is “no longer allowed for the stated purpose”.1
The final is stricter than the draft of 12 November 2024. The draft deprecated only algorithms at 112 bits of classical security strength and left stronger keys acceptable until 2035.3 The final’s tables apply “Deprecated after 2030” to every quantum-vulnerable signature and key-establishment family at 112 bits or more, so a larger RSA or elliptic-curve key no longer extends the schedule.1
Federal high-value systems, national-security buyers and contractors all face dates years before NIST’s 2035 cut-off
Post-quantum dates in US federal policy and NIST guidance, June 2026 to December 2035
| Date | Event | Detail |
|---|---|---|
| Jun 2026 | Executive Order 14412 | Sets the 2030 and 2031 deadlines for federal high-value assets and high-impact systems; OMB M-26-15 follows on 24 June |
| Oct 2026 | NIST IR 8547 final | Quantum-vulnerable public-key algorithms deprecated after 2030, disallowed after 2035 |
| Dec 2026 | FAR proposed rule due | Within 180 days of the order: a rule requiring covered contractors to comply with FIPS, including post-quantum, by 31 December 2030 |
| Jan 2027 | CNSA 2.0 for new NSS buys | New National Security System acquisitions to be CNSA 2.0 compliant unless otherwise noted |
| Jan 2030 | TLS 1.3 required | Federal agencies must support TLS 1.3 or a successor no later than this date |
| Dec 2030 | Post-quantum key establishment | Federal high-value assets and high-impact systems; NSA expects equipment that cannot support CNSA 2.0 phased out |
| Dec 2031 | Post-quantum signatures | Federal high-value assets and high-impact systems; CNSA 2.0 algorithms mandated for national security systems |
| Dec 2035 | Disallowed after 2035 | Quantum-vulnerable public-key algorithms disallowed under NIST standards; the 2035 federal goal of NSM-10 |
Source: Executive Order 14412; OMB M-26-15; NIST IR 8547 (final); NSA CNSA 2.0 FAQ v2.1; FAR date is TechBag arithmetic (180 days after 22 June 2026) · as of 2 November 2026
For the most sensitive federal systems the 2030 window does not apply. Executive Order 14412 of 22 June 2026 requires high-value assets and high-impact systems to move key establishment to post-quantum algorithms by 31 December 2030 and signatures by 31 December 2031.4 OMB’s memorandum M-26-15 of 24 June adds TLS 1.3 support by 2 January 2030 and tells agencies to align their plans with IR 8547.5 The order also gave the FAR Council 180 days to propose a rule requiring covered contractors to comply with FIPS, “including all applicable FIPS incorporating PQC compliant algorithms”, by 31 December 2030.4 National security systems are excluded from the order and follow NSA: from 1 January 2027 new acquisitions must meet CNSA 2.0 “unless otherwise noted”.7
IR 8547 lists six quantum-vulnerable families across signatures and key establishment, all on the same schedule. For key establishment, “ML-KEM is the only approved post-quantum key-establishment scheme based on public-key cryptography”, and NIST recommends ML-KEM-768 as the default parameter set. HQC has been selected as an alternative; its FIPS is still being written.1 FIPS 203, 204 and 205 were published on 13 August 2024.617
Every RSA and elliptic-curve scheme in NIST’s standards has a replacement; symmetric encryption largely stays
Quantum-vulnerable algorithms in NIST standards and their approved replacements, NIST IR 8547 final, October 2026
Quantum-vulnerable: deprecated after 2030, disallowed after 2035
Approved replacements
Source: NIST IR 8547 (final), Tables 2 to 5 and section 4.1.3 · as of 27 October 2026
Two details matter for products. Hybrid key exchange, which runs a classical and a post-quantum algorithm together, can still be FIPS 140-validated “as long as at least one of the inputs” comes from an approved scheme. And symmetric algorithms at the 112-bit level will be disallowed in 2030, with SHA-1 disallowed in FIPS-validated modules after 2030.1 NSA asks more of national security systems: ML-KEM-1024 and ML-DSA-87 at every classification level, and it “will not require” hybrid products.7
The UK’s National Cyber Security Centre published three milestones on 20 March 2025: discovery and an initial plan by 2028, the highest-priority migrations by 2031, and migration of “all your systems, services and products” by 2035.8 The EU’s coordinated roadmap, published on 23 June 2025, asks member states to start national transitions by 31 December 2026, finish high-risk use cases by 31 December 2030 and medium-risk ones by 31 December 2035.9 India’s task-force roadmap is the fastest of these, and it is a recommendation, not a mandate.15
Four roadmaps end in 2035; India’s recommended timetable for critical sectors ends six years sooner
Years from the end of 2026 to each published milestone, by roadmap; a milestone given as a year is placed at that year’s end
| To first milestone | Then to second | Then to final | |
|---|---|---|---|
| UK, all organisations | 2 | 3 | 4 |
| EU member states | 0 | 4 | 5 |
| US federal high-value systems | 4 | 1 | 4 |
| US national security systems | 0 | 5 | 4 |
| India, enterprises (recommended) | 2 | 2 | 3 |
| India, critical sectors (recommended) | 1 | 1 | 1 |
Source: UK NCSC; EU PQC Workstream roadmap v1.1; Executive Order 14412 and OMB M-26-15; NSA CNSA 2.0 FAQ v2.1; DST Quantum-Safe Ecosystem in India; TechBag count · as of 2 November 2026 · India's milestones are recommendations; the national security row starts on 1 January 2027, shown as zero
On the public web the client side moved before the rules did. Google said on 13 September 2024 that Chrome 131 would switch its hybrid post-quantum key exchange from the pre-standard Kyber to ML-KEM.13 Apple says TLS connections in iOS 26, iPadOS 26, macOS Tahoe 26 and visionOS 26 “will automatically advertise support” for hybrid key exchange.14 Cloudflare measured about two per cent of its TLS 1.3 connections as post-quantum in March 2024, over 65% of human traffic in April 2026, and over two-thirds of browser traffic in June 2026.121118
About 70 in 100 browser connections to Cloudflare use hybrid ML-KEM; about 15 in 100 of the servers behind it do
Share using hybrid ML-KEM key agreement on each leg of a Cloudflare connection, as published 29 September 2026
70%
Browser traffic reaching Cloudflare
visitor to Cloudflare
15%
Origin servers Cloudflare connects to
Cloudflare to the customer's server
Source: Cloudflare, Is your domain using post-quantum encryption? (Cloudflare Radar) · as of 29 September 2026 · Cloudflare's own measurement of its network; it does not describe the rest of the internet
The gap is the part a buyer controls. Cloudflare reports roughly 70% of browser-generated traffic reaching it protected with hybrid ML-KEM, but “just about 15% of origins that Cloudflare connects to use hybrid ML-KEM”. It names X25519MLKEM768 as “the only recommended algorithm for post-quantum encryption” in TLS 1.3.10 Each of those origins is a web server, load balancer or appliance someone bought. Authentication is a separate problem: Cloudflare’s own target for post-quantum security, “including, crucially, post-quantum authentication”, is 2029.11
IR 8547 tells organisations to begin “with cryptographic inventories, migration planning, and early deployments of PQC-enabled products and services”, and says hardware vendors should start integrating post-quantum support “immediately” because of long design and fabrication lead times.1 For a buyer that becomes a set of questions to put in writing at the next purchase or renewal. They are questions, not a test any named product passes or fails.
Ask for the inventory first, then a dated answer on key exchange, signatures, hardware and agility
Questions for vendors of software, appliances and cloud services that use public-key cryptography, November 2026
Step 1 of 6
1. Where is RSA or elliptic-curve used?
Ask for a cryptographic bill of materials (CBOM): algorithms, key sizes, protocols and libraries, per component.
Source: NIST IR 8547 (final); OMB M-26-15; CERT-In BOM guidelines v2.0, section 8.5; TechBag analysis · as of 2 November 2026
The last question has an Indian source. CERT-In’s bill-of-materials guidelines suggest “tiered vendor Post-Quantum Cryptography (PQC) contractual requirements”, with quarterly migration progress reports carrying executive attestation and a quantum-safe proof of concept before contract renewal.16 For cloud services, OMB tells US agencies to agree post-quantum responsibilities with providers “within the shared responsibility model”, a useful clause for any buyer.5 Our encryption and rights management guide lists the key-management and HSM products to put these questions to.
Sources
Related on TechBag: the encryption and rights management guide compares the key-management and HSM products; our encryption analysis covers what the 2024 draft meant for keys and HSMs; and the certificate-lifetime piece covers the TLS renewals that post-quantum certificates will ride on.
Disclosure. TechBag is a software reseller: it earns a margin when a reader buys through it, including from vendors named here. The analysis is TechBag’s own; no vendor paid for or saw this piece before publication. Facts are labelled reported, TechBag count or our view.