Detecting Malicious Agent Skills in the Wild using Attention
Explorar
Noticias de IA
1359 elementos — filtrados, clasificados y sin duplicados
Safety in Self-Evolving LLM Agent Systems: Threats, Amplification, and Case Studies
From CVE to CWE: Syscall-Based HIDS Generalisation
LambdaMark: Semantic Audio Watermarking for Robustness and Radioactivity
Local LLM Agents as Vulnerable Runtimes:A Source-Code Audit of the Agent Runtime Layer
Whose Agent Are You? Multi-Layer Fingerprinting and Attribution of Autonomous Web Agents
Signals in the Noise: Open Source Intelligence (OSINT) for AI Loss of Control Detection
AgentRiskBOM: A Risk-Scoping Security Bill of Materials for Agentic AI Systems
OpenAI launches new initiative to help find and patch open-source bugs
Red-Teaming after Mythos — Zico Kolter & Matt Fredrikson, Gray Swan
Meta Exposed Data Internally From Its Controversial Employee-Tracking Program
Tesla in autopilot crashed into Texas home, killing one
TooBad: Backdoor Diffusion Models with Ultra-Low Poison Rate and Imperceptible Trigger
The Watermark Shortcut: How Provenance Marking Sabotages Audio Deepfake Detection
Read this before you vibe-code another app
World Cup Scams Are Getting Harder to Spot
Intent-Governed Tool Authorization for AI Agents
SingGuard: A Policy-Adaptive Multimodal LLM Guardrail with Dynamic Reasoning
From CVE to CWE: Syscall-Based HIDS Generalisation
Detecting and Understanding Vulnerabilities in Fully Homomorphic Encryption Frameworks
e2e-assure introduces Cumulo, the U.K.’s only sovereign, AI-driven, zero-day SOC platform…
Analyzing Defensive Misdirection Against Model-Guided Automated Attacks on Agentic AI Sys…
LLM agent safety, multi-turn red-teaming, jailbreak benchmarks, adversarial robustness, s…
Multi-View Decompilation for LLM-Based Malware Classification
Beyond the benchmark: Advancing security at AI speed
The hacker sent by Anthropic to calm the government's nerves about AI safety
AI is accelerating cyberattacks — here’s how to stay ahead
"Dangerous" AI models are coming no matter what
Lifecycle-Aware Dynamic Analysis for Secure ML Model Execution
Membership Inference Attacks against Large Audio Language Models