Secure Coding Drift in LLM-Assisted Post-Quantum Cryptography Development: A Gamified Fix

2026-06-19T07:23:31Ze30d5d4a603902a2ca23a3629da091754651b6213842d14f53eb7c90fc569759
FHELLM-securityLoRARAGReDoShardware-aware-mitigationhubnessinference-safetymodel-backdoorplatform-dependentpost-quantum-cryptographyprivacy-preservingprompt-injectionregexretrieval-augmented-generationsecure-codingside-channelspeculative-inferencesupply-chaintiming-attackstwo-party-computationvector-poisoning

What happened

This collection of papers highlights multiple emergent attack surfaces and defenses at the intersection of LLM deployments, cryptography, and system security. Key risks: FloatDoor demonstrates a novel, platform-triggered, input-independent backdoor class (realized with lightweight LoRA adapters) that evades pre‑serving audits and can induce exploitable code vulnerabilities on specific hardware/software targets; Secure Coding Drift warns that sustained reliance on LLM-generated code produces longitudinal degradation of secure PQC implementations; PUFFERDOS produces practical ReDoS attack inputs

Why it matters

A reviewed impact interpretation has not been published for this record.

Evidence and limitations

Source ID
arxiv_cs_cr
Record identifier
e30d5d4a603902a2ca23a3629da091754651b6213842d14f53eb7c90fc569759
Enrichment time
2026-06-19T07:23:31Z
AI-assisted enrichment
Yes

This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.