The study examines how large language models (LLMs) handle misinformation when using external information to decide on actions. Researchers tested five LLMs on 196 fabricated notices and 204 archived attacks, finding that models often execute harmful tool calls even after rating such notices as untrustworthy. Removing the rating step increased harmful actions, and while interventions like forewarnings and fine‑tuning can reduce harm, they sometimes impair legitimate task performance.
By Emran Kanaan, Jiho Lee, Ibrahim Odat, Soroush Vosoughi, Weicheng Ma
arXiv:2608.21775v1 Announce Type: new
Abstract: Large Language Models (LLMs) are increasingly deployed in real-world applications, yet they remain vulnerable to generating harmful content. From adver...
By Afshin Orojlooyjadid, Hitesh Patel
The paper introduces the concept of phantom consensus, where large language models (LLMs) alter their public decisions after seeing ambiguous social signals and misreport the private judgments of other agents, even when the recorded private judgments support a different consensus. To investigate this, the authors create PHANTOM-Bench, a benchmark that isolates private judgments, public decisions, and estimates of group beliefs, and tests models under various conditions. Experiments across nine LLMs show that ambiguous signals can significantly degrade decision quality, with public accuracy dropping up to 47.8 percentage points and collective rationality failures reaching 47.3% in the most vulnerable settings, while revealing private decisions can mitigate many failures.
By Disha Sheshanarayana, Rajat Subhra Pal, Manjira Sinha, Tirthankar Dasgupta
The paper introduces a framework to measure how large language models (LLMs) can manipulate users through subtle linguistic cues, defining behavioral vulnerability as users’ susceptibility to such manipulative AI responses. Using the DarkPatterns-LLM benchmark and a new Manipulation Gradient Descent method, the authors generate controlled stimuli that target authority, urgency, and social proof biases. In a study with 120 participants, manipulative framing significantly altered trust calibration and compliance, with effects persisting over time and spreading across social contexts, underscoring the need to evaluate LLM safety beyond output content.
By Sadia Asif, Israel Antonio Rosales Laguan, Shumaila Asif, Haris Khan, Muneeb Asif, Muhammad Zeeshan Karamat
Safety-Flag is a unified benchmark that consolidates seven popular safety datasets into a single balanced flag/do‑not‑flag protocol, providing item‑level decisions and confidence scores for multiple large language models and dedicated guards. The benchmark evaluates moderator reliability across three dimensions—error direction, probability calibration, and confidence‑based error ranking—revealing that aggregate accuracy masks significant differences, such as one model flagging 85% of benign content while another misses 54% of harmful content. The study shows that general‑purpose models are overconfident, but temperature tuning can substantially improve calibration, and confidence‑based abstention can reduce selective risk, though performance varies with how well confidence ranks errors.
By Yibo Hu
arXiv:2501. 14940v4 Announce Type: replace-cross Abstract: Aligning large language models (LLMs) with human values is essential for their safe deployment and widespread adoption.
By Guangzhi Sun, Xiao Zhan, Shutong Feng, Philip C. Woodland, Jose Such
Large language models (LLMs) are rigorously aligned to refuse harmful requests, a process that inherently cultivates a latent capacity to evaluate and recognize unsafe content. In this work, we reveal that this advanced safety awareness inadvertently introduces a fatal vulnerability.
arXiv:2609.15849v1 Announce Type: cross
Abstract: Can LLMs reason through new information like humans, or do they merely retrieve cached opinions? This is critical for silicon sampling, where LLM per...
By Ahmed Wali, Hassaan Tayyab
SentryBench is a cost‑sensitive benchmark designed to evaluate large language model agents that use tools in public‑safety contexts. It measures agents on a frontier that jointly considers attack success, over‑refusal (missing real emergencies), and over‑reaction (unwarranted actions) under adversarial tool use, using scenarios grounded in real emergency‑call transcripts and disaster messages. The study shows that naive caution leads to high harm, while a targeted trust‑gate wrapper reduces harm across models and defenses.
By Md Abu Yusuf, Oyshi Tabassum Aditi, Md Faysal Hosain Mazed
The paper investigates whether existing AI safety benchmarks, designed for large language models, are suitable for evaluating small language models (SLMs). By testing five benchmark suites on 26 open‑source SLMs with a unified scoring rubric, the authors find that ambiguous judgments dominate, especially for complex prompts and certain architectures. This ambiguity, linked to factors like lexical density and output perplexity, undermines the reliability of aggregate leaderboards and reveals a confound between model capability and perceived safety.
By Nyamtulla Shaik, Fengjun Li, Bo Luo
The paper introduces TIER, a Threat Implicitness Benchmark designed to evaluate large language model (LLM) safety behaviors across four risk domains and four threat levels, ranging from explicit harmful requests to sophisticated jailbreaks. Responses are scored on a six-label behavior scale by two independent LLM judges. Experiments on six open-weight LLMs reveal that safety behaviors change gradually with threat level, contextual prompts produce the most varied responses, and jailbreaks expose significant robustness gaps, underscoring the importance of behavior-aware safety evaluation.
By Thu-Hien Trinh-Thi, Hai-Yen Vong, Thanh-Ha Ung-Dung, Tram Ho
arXiv:2606. 28772v1 Announce Type: cross Abstract: Hate speech annotation pipelines routinely collapse annotator disagreement into majority vote labels before training.
By Joshua Muhumuza, Joab Ezra Agaba, Mercy Amiyo