Here's your daily roundup of the most relevant AI and ML news for May 28, 2026. Today's digest includes 2 security-focused stories. We're also covering 6 research developments. Click through to read the full articles from our curated sources.
Security & Safety
1. Malicious npm Package Stole Files From Claude AI User Directory via GitHub
Cybersecurity researchers have discovered a new malicious package on the npm registry that comes with information stealing capabilities.
According to OX Security, the package, named "mouse5212-super-formatter," is designed to upload files from "/mnt/user-data," a dedicated directory used by Anth...
Source: The Hacker News (Security) | 22 hours ago
2. GlassWorm Malware Takedown Disrupts Developer Supply Chain Attack Infrastructure
CrowdStrike, in partnership with Google and the Shadowserver Foundation, has announced the simultaneous disruption of all command-and-control (C2) channels associated with GlassWorm, a persistent software chain campaign targeting software developers through malicious packages and extensions.
"Si...
Source: The Hacker News (Security) | 1 day ago
Research & Papers
3. Disentangling Adversarial Prompts: A Semantic-Graph Defense for Robust LLM Security
arXiv:2605.27823v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly vulnerable to adversarial prompts that exploit semantic ambiguities to bypass safety mechanisms, resulting in harmful or inappropriate outputs. Such attacks, including jailbreaking and prompt injectio...
Source: arXiv - AI | 10 hours ago
4. Adversarial Fine-tuning of Compressed Neural Networks for Joint Improvement of Robustness and Efficiency
arXiv:2403.09441v2 Announce Type: replace Abstract: As deep learning (DL) models are increasingly being integrated into our everyday lives, ensuring their safety by making them robust against adversarial attacks has become increasingly critical. DL models have been found to be susceptible to adv...
Source: arXiv - Machine Learning | 10 hours ago
5. AdvJudge-Zero: Binary Decision Flips in LLM-as-a-Judge via Adversarial Control Tokens
arXiv:2512.17375v2 Announce Type: replace Abstract: LLM-as-a-Judge systems supply the reward signal in modern RLHF and RLVR pipelines, but their binary verdict reduces to a single linear readout F_gap on one hidden state. We show this readout is shallow enough that short, low-perplexity tokens f...
Source: arXiv - Machine Learning | 10 hours ago
6. When Think-with-Image Meets Safety: What Determines Multimodal Jailbreak Robustness?
arXiv:2605.27932v1 Announce Type: cross Abstract: Think-with-image reasoning is emerging as a new inference paradigm for large vision-language models, but its safety implications remain poorly understood. Existing systems already span multiple process designs, including direct response generatio...
Source: arXiv - Machine Learning | 10 hours ago
7. MIRAGE: Context-Aware Prompt Injection against Mobile GUI Agents via User-Generated Content
arXiv:2605.28116v1 Announce Type: cross Abstract: Mobile graphical user interface (GUI) agents driven by vision-language models (VLMs) perceive the screen as rendered pixels and choose actions from what they see, so they cannot reliably separate trusted interface elements from user-generated con...
Source: arXiv - AI | 10 hours ago
8. LLM Zeroth-Order Fine-Tuning is an Inference Workload
arXiv:2605.28760v1 Announce Type: new Abstract: Zeroth-order (ZO) fine-tuning is attractive for large language models because it replaces backpropagation with forward objective evaluations. Existing implementations nevertheless execute ZO algorithms inside conventional training loops, even thoug...
Source: arXiv - Machine Learning | 10 hours ago
About This Digest
This digest is automatically curated from leading AI and tech news sources, filtered for relevance to AI security and the ML ecosystem. Stories are scored and ranked based on their relevance to model security, supply chain safety, and the broader AI landscape.
Want to see how your favorite models score on security? Check our model dashboard for trust scores on the top 500 HuggingFace models.