🛡️ AI Security - CWhiting · Scour

Oracle Poisoning: Corrupting Knowledge Graphs to Weaponise AI Agent Reasoning 🤖Artificial Intelligence

arxiv.org·2d·Hacker News

How are you handling prompt injection across multi-step agent workflows? 🧠Context Engineering

msukhareva.substack.com·6d·Substack

The Other Half of AI Safety 🛡️AI Safety

personalaisafety.com·17h·Hacker News

A pragmatic threat model for AI coding agents, with controls you can ship today 📋AGENTS.md

dev.to·9h·DEV

AI security is broken at runtime: Most enterprises don’t realize it yet 🤨AI Criticism

·1d

AI safety tests have a new problem: Models are now faking their own reasoning traces 🤨AI Criticism

the-decoder.com

·6d

Mythos Preview is the first AI model to complete both of AISI's cyber ranges, which measure models' cyberattack capabilities; GPT-5.5 solved only one of them (A... 🤖AI News

techmeme.com·20h

Sweet Security Launches Agentic AI Red Teaming to Counter ‘Mythos Moment’ 💻Coding Agents

securityweek.com·1d

AI cyber capability is speeding past earlier projections 📊AI Benchmarks

helpnetsecurity.com·8h

How to verify AI-discovered vulnerabilities aren't just training data echoes 🛡️Exploit Mitigations

dev.to·2d·DEV

My Agentic Trust Issues: From Prompt Injection to Supply-Chain Compromise on gemini-cli 🐛Bug Bounty

pillar.security·5d·Lobsters

Anthropic forms $200 million partnership with the Gates Foundation 🤖Anthropic Claude API

anthropic.com·2h·Hacker News

MCPSafe — free security scanner for MCP servers (looking for FPR feedback) · modelcontextprotocol 🛡️macOS Security

github.com·21h·DEV

The Role of AI Safety Standards in Modern MLOps 🛡️AI Safety

mlops.community·1d

How to prevent AI’s 2008 moment (We're hiring) 🛡️AI Safety

lesswrong.com·6d

I Want to Be a von Neumann Probe: Why We Need to Fix AI Safety 🛡️AI Safety

justinldew.substack.com·1d·Substack

How a Morse Code Message Hacked Grok: Lessons in AI Security for Developers 💉Prompt Injection

dev.to·6d·DEV

Hugging Face hosted malicious software masquerading as OpenAI release 🛡️OWASP

artificialintelligence-news.com·2d

New Claude Mythos becomes the first AI model to clear all cyberattack simulations from Britain's AI safety agency 🤖Anthropic Claude

the-decoder.com

·6h

1970 exploitable findings later. 💣Binary Exploitation

hackernoon.com·2d

Log in to enable infinite scrolling