Engineering the Post-Quantum Transition at Scale: Protocol Integration, Library Defaults, and the In-Between Era of Internet Security (2025)

Abstract

Background: The August 2024 finalization of FIPS 203, 204, and 205 established the cryptographic core of the post-quantum transition; 2025 was the year that core was engineered into the default path of the Internet . Within twelve months, ML-KEM-based hybrid key exchange became a default, zero-configuration behavior across the major browser engines, a mainstream language runtime (Go 1.24's crypto/mlkem standard-library package), and — decisively for server operators — OpenSSL 3.5.0 (April 2025, LTS), whose default TLS keyshares offer X25519MLKEM768. By late October 2025, more than half of human-initiated traffic observed at a major CDN was post-quantum protected. Objective: This paper characterizes the 2025 state of the post-quantum transition as an engineering achievement, and identifies what the achievement does not yet cover. Methods: Following a design-analytical methodology, we (i) verify the 2025 default-enablement sequence against primary release documentation (OpenSSL, Go, browser engines, platform vendors, CDN telemetry), (ii) analyze the standards pipeline as it stood at year-end (the HQC selection of March 2025; the pending FN-DSA draft; the IR 8547 deprecation schedule), and (iii) construct a residual-risk register mapping six risk classes against six mitigation classes. Results: The deployment achievement is real but asymmetric: confidentiality (key establishment) has moved to post-quantum hybrids at scale, while authentication remains entirely classical — public post-quantum certificates are not expected to be widely issued before 2026 and broadly trusted later. We term this the in-between era and show that it closes the harvest-now-decrypt-later window for transit confidentiality while leaving an open window for identity forgery and server impersonation. The residual-risk register identifies implementation maturity (the KyberSlash-class timing-defect lesson), certificate and PKI latency, and the non-upgradeable long-tail client population as the dominant open risks. Conclusions: The engineering problem of 2025 was enabling post-quantum key exchange by default; the engineering problem of 2026–2027 is post-quantum authentication on the same schedule — and organizations should plan PKI re-issuance, validation pipelines, and legacy-exception registers against that window now.