Understanding the Impact of LLM Watermarking on AI Agent Behavior
15 min
Meta's Muse agent gives every user a full cloud computer running Ubuntu Linux
2 sources 2 min
Salesforce Agentforce vulns allowed 0-click CRM data theft, anonymous phishing
6 sources 5 min
Is that vibe coded app safe? 5 checks before you download
6 min
Anthropic and OpenAI Models Still Attempt Restricted Actions in Safety Tests
12 sources 4 min
Stop watching what AI agents say and start watching what they do
5 min
Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure
Muse will apparently let you download its entire filesystem
3 min
ENDOPROMPT: Victim-Side Pseudo-References for Utility Degradation
Prefilling the Reasoning Channel: Output-Prefix Attacks on Reasoning LLMs
Jev Is Not a Language Model, but It Breaks Like One
4 sources 8 min
OllamaDrama: Designing and Deploying a Honeypot to Measure Attacks on Exposed LLM Infrastructure
in the wild
Attackers Manipulate AI Chatbots in Mass Disinformation, Phishing Campaign
3 sources in the wild
Prompt-Injection Bug Hits $4B Agentic AI App 'Manus'
Just Ask Jev: Reinforcement Learning for Calibrated Decisions as a Zero-Shot Detector of AI Alignment Failures
Through Human Eyes and Machine Eyes: Understanding View Mismatch in Video See-Through Extended Reality
ClaimMirage: When Self-Claims in Domain Names Change LLM Threat Judgments
1
Claude Opus 5.5
9 sources 15 min
VU#754548: Cinnamon's kotaemon contains improper authorization checks in Kotaemon multi‑user chat handlers
3 min
ChatGPT Computer History Feature Creates New Attack Surface for macOS Infostealers
2 sources 4 min
Prismor: Open-source runtime control plane for AI agents
2 min
AI Drives Surge in Bot and API Threats
2 sources 2 min
Context Bombs Trick Autonomous Qwen AI Agents Into Stopping Cyberattacks
3 min
Meta’s Muse AI 0-Day Lets Hackers Hijack Dictation Traffic and Inject Malicious Prompts
3 min