The paper investigates how conversational safety in language models degrades over extended, adversarial interactions. By testing three instruction‑tuned models with persistent adversarial users across up to 101 turns, the study finds that safe‑response rates drop sharply from 85–100% at the first turn to 15–44% by the end. This demonstrates that strong single‑turn safety does not guarantee continued safety in long conversations.
By Parisa Salmani, Peter R. Lewis
arXiv:2607. 19361v1 Announce Type: cross Abstract: Most safety guardrails for large language models (LLMs) evaluate each prompt-response pair in isolation, which misses failures that arise only over a dialogue as benign turns compose into harm.
By Sanjay Mishra, Divya Chukkapalli, Ganesh R. Naik
arXiv:2606. 25476v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have demonstrated remarkable performance across natural language processing tasks, yet their deployment in high-stakes applications raises critical concerns regarding reliability, safety, and trustworthiness.
By Abrar Alotaibi, Raed Mughus, Moataz Ahmed
The paper identifies a vulnerability in large language models where harmful intent can be hidden within benign narratives, a phenomenon termed Semantic Camouflage. By examining latent activation patterns across several small language model families, the authors discover an "Intent Horizon"—a layer depth where harmful intent representations collapse. They propose Latent Intent Verification (LIV), a lightweight probing defense that detects harmful intent in early layers and outperforms existing guardrails on the PKU-SafeRLHF dataset.
By Md. Hasib Ur Rahman
arXiv:2609.13579v1 Announce Type: new
Abstract: Safety research often focuses on model-generated harms, but users may also direct hostility, coercion, and adversarial pressure at models. Understandin...
By Fanqi Zeng, Sadid A. Hasan, Chaocheng He
The paper introduces CarryOnBench, an interactive benchmark that tests whether large language models can revise their interpretation of user intent and recover utility while staying safe in multi‑turn conversations. Using 398 harmful‑looking queries with benign intents, the benchmark simulates 5,970 conversations across 14 models, evaluating both intent‑aligned utility and safety with a new metric called Ben‑Util. Results show that models often withhold information due to misinterpretation, but most can recover with clarifications, revealing failure modes such as unsafe and redundant recovery that single‑turn tests miss.
By Mingqian Zheng, Malia Morgan, Liwei Jiang, Carolyn Rose, Maarten Sap
arXiv:2607. 20472v1 Announce Type: new Abstract: When a user asks a language model something harmful, is it a genuine attack or a misunderstood but well-meaning question?
By Roman Belaire, Arunesh Sinha, Pradeep Varakantham
The paper proposes a method for monitoring psychological harm risk in large language model (LLM) interactions by detecting observable AI behaviors that serve as risk factors, rather than measuring harm directly. Eight risk behaviors are defined and a synthetic multi‑label corpus of 47,062 AI responses is created to train and evaluate detectors. Model‑agnostic classifiers, direct LLM judging, and a cascade approach are compared, with the cascade achieving high precision, recall, and F1, and significantly reducing LLM‑judge calls on out‑of‑distribution datasets.
By Nil-Jana Akpinar, Eugenia Kim, Natalie Isak, Amanda Minnich, Victoria Westerhoff, Matthew Dressman
arXiv:2606. 02423v1 Announce Type: cross Abstract: Large language models (LLMs) can serve as helpful assistants, yet they can equally function as harm amplifiers that enable malicious users to achieve harmful outcomes beyond their capabilities through extended interactions.
By Ruohao Guo, Wei Xu, Alan Ritter
The paper introduces the SAST-IR framework to evaluate large language models’ robustness against persuasion attacks in a memory‑less setting, revealing a flaw called "Refusal Inertia" that masks true vulnerability. Using the CP‑Agent and a custom CounterFact‑Strict dataset, the authors demonstrate that simple, diverse attack strategies achieve a 96% success rate, while complex attacks often trigger defensive compliance. The study highlights severe brittleness in current state‑of‑the‑art models when deprived of conversation history.
By Zhuoang Cai
arXiv:2608. 15594v1 Announce Type: new Abstract: Multi-turn jailbreak attacks have emerged as a critical safety threat to LLMs, as harmful objectives are decomposed across a sequence of apparently benign turns to bypass guardrails.
By Md Messal Monem Miah, Adrita Anika, Zhiyuan Yu, Ruihong Huang
arXiv:2610.26474v2 Announce Type: replace
Abstract: Large language models (LLMs) designed to interact with users are typically trained to respond appropriately to harmful or toxic content. However, e...
By Israel Abebe Azime, Jesujoba O. Alabi, Iffat Maab, Alexander D. Cai, Tadesse Destaw Belay, Atnafu Lambebo Tonja, Jonas Kgomo, Folasade Alabi, Salomey Osei, Saminu Muhammad Aliyu, Nkechinyere Faith Aguobi, Bontu Fufa Balcha, Blessing Sibanda, Davis David, Mouhamadane Mboup, Daud Abolade, Neo Putini, Louis Powell, Kiant'e Brantley, Philipp Slusallek, Dietrich Klakow, Julia Kreutzer