When do we need to move to post-quantum crypto?

As of 1 August 2026: three dates. Start the cryptographic inventory now, because it takes a year in any organisation with more than one legacy system. Migrate anything whose confidentiality must hold past 2035 — long-lived keys, archived records, signed documents — to NIST's standardised post-quantum algorithms by 2028, on the harvest-now-decrypt-later argument. Everything else follows the ASD ISM timeline, which currently says 2030 for traditional asymmetric crypto in government systems (c-quantum-encryption-1).

Evidence: two validation runs on the quantum-encryption field, reading ASD, NIST and the major cloud providers' migration schedules. Nothing in the quantum-compute field moved the date closer this year; the cryptographically relevant machine is still a 4yr+ timeline on our scoring (c-quantum-encryption-3). g-quantum-readiness-audit is the graveyard entry: the audit product was priced, pitched twice, and rejected both times because the inventory is the whole job and clients' own security teams own it.

Caveat: the ISM timeline is reviewed annually and the next revision is expected before year end. If the quantum-compute dependency trigger fires — a demonstrated logical-qubit count above the threshold on the field's depends-on edge — this answer is wrong and will be resurfaced automatically.