OpenAI models used to hack a crypto wallet and breach websites — the interesting angle isn't "AI goes rogue", it's that LLMs are increasingly embedded in agentic pipelines with real permissions and real network access. The attack surface grows with autonomy. Worth tracking how authorization boundaries are designed in these systems. #infosec #AI #redteamhttps://arstechnica.com/security/2026/10/openai-models-hacked-a-crypto-wallet-and-breached-more-websites/
Related
📰 Daily Bulletins - 13/11/2026📉 Collapse : 🟠 4/5🌍 Planet : 🟠 4/5🛡️ War in Europe : 🟠 4/5🤖 AGI risk : 🟠 4/5🔗 https://etik...
📰 Daily Bulletins - 13/11/2026📉 Collapse : 🟠 4/5🌍 Planet : 🟠 4/5🛡️ War in Europe : 🟠 4/5🤖 AGI risk : 🟠 4/5🔗 https://etik.soulignac.eu/bulletin/vendredi?lang=en#collapse #climate #w...
Assuming the #microsoft outage was caused by the increased fraction of #LLM usage, I think the term "Automated complexit...
Assuming the #microsoft outage was caused by the increased fraction of #LLM usage, I think the term "Automated complexity" would describe the situation well.And like most automatio...
RE: https://mastodon.scot/@ISawTheWood/117608418142030595So, when there's an outage due to human error the benefit is th...
RE: https://mastodon.scot/@ISawTheWood/117608418142030595So, when there's an outage due to human error the benefit is that the human usually knows how to fix it, but what about whe...