The paper "Mind the Gap: Robustness Risks in PII Detection Systems" examines how modern PII detection methods—encoder-based NER, rule-based hybrid detection, and generative LLM extraction—fail when faced with realistic distribution shifts such as noisy, unstructured, and informal inputs. A stress‑test benchmark covering seven categories of natural shift reveals that each architecture degrades in distinct ways: encoder models struggle with unseen surface forms and boundary detection, rule‑based systems falter on non‑standard formats, and LLMs suffer from entity‑type confusion and generation instability. The authors argue that aggregate benchmark scores hide these deployment‑critical weaknesses and propose a hybrid detection pipeline with a QA‑driven feedback loop to iteratively mitigate risks, also releasing the benchmark for OOD‑aware evaluation.
arXiv:2606. 20502v1 Announce Type: cross Abstract: Whether LLMs scoring well on vulnerability benchmarks genuinely reason about security or merely pattern-match on contaminated data remains unresolved.
arXiv:2606. 00052v1 Announce Type: new Abstract: As Industry 4.
The paper proposes a neurosymbolic defense architecture for AI-enhanced Security Operations Centers (AI‑SOCs) that protects against indirect prompt injection via log poisoning. It combines deterministic SIEM decoders as a pre‑filter with NeMo Guardrails for semantic validation, and adds a closed‑loop telemetry system for Human‑in‑the‑Loop visibility. Experimental results mapped to the MITRE ATLAS taxonomy show the approach effectively dismantles promptware kill chains and delivers a resilient, observable defense for next‑generation AI‑SOCs.
arXiv:2609.15017v1 Announce Type: cross Abstract: Prompt-injection detectors are typically evaluated using aggregate F1 on in-distribution test data, which offers limited insight into behavior under...
arXiv:2606. 09864v1 Announce Type: cross Abstract: Key-value (KV) cache quantization is widely used to reduce Large Language Model (LLM) inference memory, yet existing evaluations solely focus on measuring perplexity and accuracy without assessing the safety impact.
arXiv:2608.20554v1 Announce Type: cross Abstract: The critical failure modes in deployed large language models (LLMs) are cross-dimensional: a model can score 99.3 in safety alignment while refusing...
InGuard introduces an inner guardrail for text-to-image generation that operates within the model’s own representations, avoiding external classifiers. It grades prompts using the text encoder’s embeddings, modifies risky embeddings with SAGE to produce safe images, and employs a latent detector to halt generation early. Evaluated on the RevGen Safety Benchmark, InGuard achieves a 97.9–98.8% safety rate across five open-weight models while reducing benign disturbances, model parameters, and denoising steps.
arXiv:2606. 26396v1 Announce Type: new Abstract: Pre-trained transformers have demonstrated remarkable generalization abilities, at times extending beyond the scope of their training data.
arXiv:2608. 10621v1 Announce Type: new Abstract: Recent research on Large Language Model (LLM) safety has widely adopted guardrails to identify unsafe LLM outputs.
arXiv:2610.07532v1 Announce Type: cross Abstract: LLMs have advanced rapidly, raising growing concerns about their safety. Recent work has proposed approaches to detect and defend against attacks inc...
arXiv:2609.14593v1 Announce Type: cross Abstract: Living-Off-the-Land (LOTL) is the dominant evasion technique of Advanced Persistent Threat (APT) actors, exploiting legitimate Windows utilities to c...
PGuard is a native, streaming guardrail for multimodal large language models that integrates safety monitoring directly into the backbone model’s inference pipeline. It uses dynamic layer gating to aggregate intermediate representations for token‑level safety detection without altering generation behavior, offering lightweight and sequence‑aware variants to balance accuracy and efficiency. Evaluations on over 30 benchmarks show state‑of‑the‑art performance with minimal runtime overhead (0.82‑2.26%) on a 122B‑parameter backbone.