White House Sets Hard Clock on Post-Quantum Migration for Federal Systems
An executive order mandates that high-value federal assets shift to post-quantum cryptography by 2030-2031. For identity infrastructure, that deadline is closer than it looks.
Key points
- The executive order requires federal agencies to migrate high-value assets to post-quantum cryptography by end of 2030, and high-impact systems by 2031.
- SAML assertions, OIDC tokens, TLS sessions, and PIV/CAC certificates are all in scope for replacement.
- NIST finalized the ML-KEM and ML-DSA standards in FIPS 203 and FIPS 204 in August 2024; vendor adoption remains early-stage.
- The harvest-now-decrypt-later threat means adversaries may already hold encrypted federal traffic they plan to decrypt once capable quantum hardware exists.
- Whether hybrid classical-plus-PQC deployments will count toward the 2030 deadline is unresolved and matters for procurement now.
The White House has signed an executive order requiring federal agencies to migrate high-value assets to post-quantum cryptography (PQC) by the end of 2030, and high-impact systems by 2031. Six years sounds comfortable. For anyone who has tried to rotate a root certificate authority across a federal identity estate, it isn't.
Where identity infrastructure runs into trouble
Most federal SSO and PKI infrastructure relies on RSA or elliptic-curve key exchange, both vulnerable to Shor's algorithm at sufficient qubit scale. That puts SAML assertions, OIDC ID tokens, TLS sessions protecting LDAP and Kerberos traffic, and the certificates underpinning PIV and CAC authentication squarely in scope. Replacing any one of them involves vendor coordination, hardware security module firmware, certificate lifecycle tooling, and identity provider support for NIST-standardized algorithms like ML-KEM (formerly KYBER) and ML-DSA (formerly DILITHIUM), finalized in FIPS 203 and FIPS 204 respectively in August 2024. As we reported when those standards landed on 23 June 2026, only 5% of security teams had a defined strategy a year after publication.
Should you worry about harvest-now-decrypt-later?
Yes, and the executive order doesn't change the underlying exposure. Nation-states with storage budgets have been collecting encrypted federal traffic for years, planning to decrypt it once capable quantum hardware exists. Authentication tokens and session material are attractive targets. A stolen, encrypted SAML response from 2024 could, in theory, be cracked in 2031 and used to reconstruct session context or inform credential-stuffing against legacy systems that haven't rotated.
MFA doesn't solve this. Hardware keys don't solve this. The problem sits in the asymmetric primitives that protect key agreement and digital signatures, not in the second factor on top.
What the order leaves open
The order doesn't resolve the hybrid transition question. NIST guidance recommends running classical and PQC algorithms in parallel while the ecosystem matures, a dual-algorithm handshake being defined in ongoing RFC work. Whether agencies can count hybrid deployments toward the 2030 deadline is unresolved. That ambiguity will shape procurement decisions made this year, not in 2029.
The plain judgement: the deadline is real, the identity layer is the hardest part, and the vendors who haven't committed to FIPS 203 and FIPS 204 support on a public roadmap are the ones worth pressing right now.
