ReCon: A Resource-Constrained Benchmark for LLM-Based Cybersecurity Compliance Across Ingestion and Retrieval Pipelines
2026-07-28T07:23:25Z•92d0bbdab5e6875fcdb3ab10d5b8394990616c5c8858396f27501d9cda946e94
LLM securityOT securityV2X securityadversarial machine learningagentic systemsautonomous vehiclescode generation securitycompliance automationconstrained devicescritical infrastructurecybersecurity researchfeature injectionfraud detectionhardware securitypost-quantum cryptographypower side channelsprompt and model attackssmart grid
What happened
This batch contains research on cybersecurity compliance benchmarking, cyber ranges for smart energy systems, post-quantum cryptography for constrained networks, LLM safety and agentic-system risks, V2X perception attacks, fraud detection, insecure LLM-generated code, and hardware power side-channel analysis. The most directly security-relevant findings include stealthy multi-attacker feature injection against collaborative autonomous-driving perception, world-model manipulation causing harmful agent actions and data exposure, high vulnerability rates in LLM-generated code under realistic risk
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_cs_cr
- Record identifier
- 92d0bbdab5e6875fcdb3ab10d5b8394990616c5c8858396f27501d9cda946e94
- Enrichment time
- 2026-07-28T07:23:25Z
- 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.