🛡️ AI Security - inarcissuss · Scour

💉Prompt Injection arXiv·

GIF: Locally Sound Geometric Information Flow Control for LLMs

🤖LLM, Agent arXiv·

Local LLM Agents as Vulnerable Runtimes:A Source-Code Audit of the Agent Runtime Layer

🤖LLM, Agent arXiv·

Securing LLM-Agent Long-Term Memory Against Poisoning: Non-Malleable, Origin-Bound Authority with Machine-Checked Guarantees

🧠LLM arXiv·

A Layered Security Framework Against Prompt Injection in RAG-Based Chatbots

💬LLM Prompting arXiv·

Detecting Malicious Agent Skills in the Wild using Attention

💬LLM Prompting arXiv·

Analyzing the Narration Gap in LLM-Solver Loops

💉Prompt Injection Simon Willison’s Weblog·

Prompt Injection as Role Confusion

Covers A Theory of Why Prompt Injection Works

Covered by 3 sources including Schneier on Security, tldr.tech

💉Prompt Injection Schneier on Security·

Interesting Paper Exploring Prompt Injection

Covers 3 stories including A Theory of Why Prompt Injection Works

🔐Cybersecurity BleepingComputer·

New macOS malware embeds fake errors to confuse AI analysis tools

Covers macOS.Gaslight | Rust Backdoor Turns Prompt Injection on the Analyst, Not the Sandbox

Discussed on Hacker News

🔐Cybersecurity The Hacker News·

New Gaslight macOS Malware Uses Prompt Injection to Disrupt AI-Assisted Analysis

Covers macOS.Gaslight | Rust Backdoor Turns Prompt Injection on the Analyst, Not the Sandbox

Covered by sh.itjust.works

🔐Cybersecurity Infosecurity Magazine·

macOS Backdoor Uses Prompt Injection to Evade AI Triage

Covers macOS.Gaslight | Rust Backdoor Turns Prompt Injection on the Analyst, Not the Sandbox

🦀Rust Malware Analysis, News and Indicators·

macOS.Gaslight | Rust Backdoor Turns Prompt Injection on the Analyst, Not the Sandbox

Covers 3 stories including Mini Shai-Hulud, Miasma, and Hades Worms Target Bioinformatics and MCP Developers via Malicious PyPI Wheels

Log in to enable infinite scrolling