GAS-Leak-LLM: An algorithm for black-box jailbreak attacks
Automatic suffix optimization as a way to test LLM robustness
Ai
AI Cybersecurity Careers 2026
Career overview of the AI Security market for 2026
AI Security
Courses and certifications in AI Security
MCPThreatHive: Automated Threat Intelligence for MCP Ecosystems
A threat discovery platform for Model Context Protocol
RLSpoofer: A Lightweight Tool for Evaluating Watermark Spoofing Robustness
An attack on watermarks as a way to spoof authenticity
DeepSight
A transition from black-box evaluation to transparent AI safety diagnostics
Technology Predictions for 2026 from IEEE
IEEE released a report covering the key technology trends expected to shape 2026.
OpenRT - An Open Framework for Red Teaming Multimodal LLMs
OpenRT is a modular and extensible environment for systematic safety evaluation of large language models
Small Language Models
Note on the document Small Language Model for AI Agents HandBook
Doublespeak
The authors present a new attack called Doublespeak: a simple attack based on hijacking the model’s internal representations in context