Ethereum may be pivoting from Poseidon to SHA/BLAKE, as binary-field SNARKs make old SNARK-friendly hash orthodoxy look dated.

New quasipolynomial distinguisher for Classic McEliece gets compared to ISD and, annoyingly, still isn't a break.

AI-assisted cryptanalysis makes “independent” checks less independent, since shared models and tooling can correlate failures.

When do signatures catch up? Google Cloud's PQC roadmap says about a year after confidentiality, if standards and clients cooperate.

Analysis of DARPA QBI's stage-gating, where Stage C can make the least demonstrated qubit plans look furthest along. Convenient.

Draft claims a polynomial-time quantum attack on DCP, with shaky reductions to LWE and possible ML-KEM fallout.

Unsigned CA cosigner certs are now a thing, as TrustAsia files for Chrome’s MTC test root store.

MAS is giving banks a 2026 heads-up and an end-of-decade deadline to stop pretending RSA lasts forever.

Fact-checks IBM’s three quantum-advantage preprints, and concludes the evidence is still too thin to move CRQC or PQC timelines.

87% say they're pursuing PQC, but only 7% have deployed at scale, so the execution gap is doing fine.

WEF's 2026 tech list mislabels PQC, and this audit walks the errors back before anyone budgets for them.

HKMA’s PQC readiness whitepaper for Hong Kong banks, with a 2.3/10 index and a 2030 migration target, because procrastination.

Essay pushing back on the 'always five years away' line, with CRQC, fault-tolerance, and PQC targets. Timelines, naturally, still wobble.

AI cracked a HAWK candidate and reduced-round AES-128, not your encryption, a survey of what was actually broken.