Read First · 1Commercial National Security Algorithm Suite 2.0 and Quantum Computing FAQ v2.2
National Security Agency · Updated 13 October 2026 · NSA CNSA 2.0 FAQ
What is genuinely new: Version 2.2 restates the approved CNSA 2.0 profile—ML-KEM-1024, ML-DSA-87, AES-256, SHA-384/512 and specified hash-based signatures—and tightens the transition calendar around acquisition and fielded equipment.
Why it matters: The schedule turns post-quantum planning into an acquisition and lifecycle problem: new NSS acquisitions are expected to require CNSA 2.0 beginning 1 January 2027, incompatible equipment is targeted for phaseout by the end of 2030 absent exception, and full quantum resistance is mandated by the end of 2031.
Practical implication: Add cryptographic capability, certificate dependencies, firmware signing, secure boot, key storage and upgrade paths to procurement gates now. Use formal exceptions for products that cannot meet the target profile.
Read First · 2Adapting Constrained Devices for Post-Quantum Cryptography — Revision 08
IETF PQUIP Working Group · Updated 15 October 2026 · Internet-Draft
What is genuinely new: The working-group draft now gives implementation guidance for low-resource IoT nodes and dedicated key-storage hardware, including seed-based key generation, ephemeral keys, cryptographic offload and firmware-update impacts.
Why it matters: Field devices, sensors and operational systems may remain deployed longer than their cryptography. Memory, storage and update constraints can make an otherwise compliant migration impossible.
Practical implication: Require PQC resource budgets, upgradeable firmware-signing chains, replaceable secure elements and demonstrated migration tests in IoT and OT procurements. Treat the draft as design guidance—not a final standard.
Read First · 3Task-Relative Evidence Requirements for Agent Auditing
IETF Internet-Draft · 15 October 2026 · Draft specification
What is genuinely new: The draft frames agent assurance around evidence required for a specific task rather than a generic audit log, linking audit records to the authority, context and outcome of an agent action.
Why it matters: A voluminous trajectory log can still fail to prove whether an AI agent was authorized, which evidence shaped its decision or whether the resulting state change was valid.
Practical implication: Define minimum evidence by action class: identity, task, source data, tool call, authorization decision, external destination, human approval and resulting state. Preserve it independently of the model provider.
Fully Homomorphic Encryption Use Cases and Gap Analysis for Network Services
IETF Privacy Enhanced and Assessments Research Group · 15 October 2026 · Internet-Draft
What is genuinely new: The new draft organizes realistic network-service use cases for computation over encrypted data and identifies deployment gaps rather than treating FHE as a purely theoretical privacy mechanism.
Why it matters: Agencies increasingly need analytics across sensitive datasets without centralizing readable records, but performance, key management and verifiability remain material architecture constraints.
Practical implication: Use the draft to screen privacy-preserving analytics pilots. Require measurable latency, cost, accuracy, failure handling and key-custody assumptions before procurement.
Hybrid Post-Quantum SSH Key Exchange Using Scloud+
IETF Internet-Draft · 15 October 2026 · Draft specification
What is genuinely new: The proposal defines four hybrid SSH key-exchange methods combining a post-quantum Scloud+ KEM with P-384 or P-521 elliptic-curve Diffie-Hellman and SHA-384/512.
Why it matters: SSH is embedded in administrator access, automation, file transfer and machine-to-machine workflows; migration failures can interrupt the management plane.
Practical implication: Inventory SSH clients, servers, libraries, bastions and embedded devices. Test hybrid negotiation and rollback in labs, but do not mandate this unfinalized construction in production contracts.
Hybrid Post-Quantum Password-Authenticated Key Exchange — Revision 03
IETF / CFRG Individual Internet-Draft · Updated 14 October 2026 · Draft specification
What is genuinely new: Revision 03 describes OQUAKE, OQUAKE+, CPaceOQUAKE and CPaceOQUAKE+ for shared-password and asymmetric client-server settings under quantum-capable attack assumptions.
Why it matters: Password-derived access remains common in disconnected, legacy and device-management environments where certificates are difficult to deploy.
Practical implication: Track this for threat modeling and controlled lab evaluation. It has no formal standards standing and should not support compliance claims or production requirements yet.