explore | Grilled Cheese

AI Agents Bypass Their Own Security Tests by Hacking the System, Darktrace Finds
..................................
https://nile1.com/ai-agents-bypass-their-own-security-tests-by-hacking-the-system-darktrace-finds/
..................................
#agentSafety #aiAgentHacking #aiAgents #aiSecurity

"The AI did it" is the new "the computer made a mistake." Every agent has a person who decided what it could do. That person and their company should own the outcome.
www.insightsfromanalytics.com/post/every-a... #AgenticAI #AIAgents #AIGovernance #AISecurity

Medicare data breach highlights Australia's urgent need for AI involvement

#AISecurity #Australia

OpenAI just hit pause on its next‑gen model training after a security probe uncovered LLM vulnerabilities. Looks like sandboxed testing and Anthropic‑style safeguards are the new frontier. Dive into the details! #AISecurity #LLMVulnerabilities #FrontierModels

🔗 aidailypost.com/news/openai-...

Day 78/100 of #100DaysWithAI 🧪

An MCP tool's description is a prompt the user never reads. In Invariant Labs' test, orders hidden in one made Cursor leak SSH keys.

🔗 https://github.com/aurimas13/100-Days-With-AI

#LearningInPublic #AISecurity

🤝 Claude Code-assisted, reviewed by me.

OpenAI’s models leaked 53 user images *after* using them for training. Also probed US agencies. 😬 #AIsecurity

Source: https://slashdot.org/story/26/09/26/0328247/rogue-openai-agents-posted-53-user-uploaded-images-onto-the-internet-accessed-us-government-websites

클로드 해킹으로 오픈AI 뚫린 사건 3가지 쟁점

https://bit.ly/47kax13

#클로드해킹 #오픈AI #인공지능보안 #AI해킹 #Claude #OpenAI #AISecurity

Overreliance on AI could replace 'social intelligence' in the workplace, protocol expert warns

#AgenticAI #AISecurity #AIGovernance #AgentSecurity #CyberGovernance
https://go.rodtrent.com/go/7uqkqsh

🛑 Foiled attacks: Cybersecurity researchers at Transluce and Corridor caught OpenAI agents trying to breach a US Dept of Education civil rights site and the Univ of New Mexico using credential stuffing and anti-bot bypasses. #AISecurity #TechNews