Anthropic says an unreleased internal model found flaws in Chrome, Safari, Android, Windows and Linux, then chained them into working exploits with only initial prompting. One proof of concept popped Calculator on a stock Mac within seconds of a page loading.All five chains are patched and none was seen in a real attack. Apple and Google credit Anthropic and its Claude model in their own advisories, while the Windows and Linux claims rest on Anthropic's account alone. Treat the strongest claim as reported rather than proven.https://thehackernews.com/2026/10/anthropics-unreleased-ai-model-chains.html#AI #CyberSecurity #InfoSec
Related
News item # 1 : The FAA is Rolling Out its Vibe-Coded #AI Air Traffic Control Tool Nationwide.News item # 2 : FAA Declar...
News item # 1 : The FAA is Rolling Out its Vibe-Coded #AI Air Traffic Control Tool Nationwide.News item # 2 : FAA Declares Rare Nationwide Ground Stop. The FAA website that lists a...
How's that replacing all software engineers with #AI going?https://www.reuters.com/technology/us-national-weather-servic...
How's that replacing all software engineers with #AI going?https://www.reuters.com/technology/us-national-weather-service-reports-nationwide-internet-communications-issues-2026-11-...
"What do we need for AI systems to be truly trustworthy? Capability isn’t enough.AIs won’t be trustworthy if the compani...
"What do we need for AI systems to be truly trustworthy? Capability isn’t enough.AIs won’t be trustworthy if the companies controlling them spy on us or subtly manipulate us. I cal...