This week, we’re seeing a pattern emerge across the industry when it comes to AI. Anthropic is dealing with restrictions on its newest model put in place by the US government. Researchers tricked AI agents into handing over credentials and customer data and Microsoft unfortunately is once again cleaning up after a supply chain compromise. AI is obviously getting more capable but it is also creating entirely new security problems that we all have to come to terms with and I’ll be talking about this and much more on this week’s episode.
Episode 444 is available on Soundcloud:
On YouTube:
If you’d like to play the Podcast on a different podcast service such as Apple Podcasts or Spotify. You can find this episode at 5BytesPodcast.com
Reference Links:
News:
Anthropic Forced to Pull Fable 5 Model
https://www.bbc.com/news/articles/c932g3v3e13o
June Windows Update Failing to Install
Fable 5 Was Being Used in Some Interesting Ways
https://www.anthropic.com/news/claude-fable-5-mythos-5
Xbox CEO Speaks Candidly About AI
https://www.tomsguide.com/ai/xbox-ceo-just-scrapped-copilot-ai-for-consoles-and-i-couldnt-be-happier
5th Chrome Zero Day Vulnerability This Year
OpenClaw Can be Tricked by Phishing
ServiceNow Customer Instances Breached
Microsoft Repo Hit Again
NPM Getting Some Added Protections
Ivanti Sentry Flaw Being Exploited
Controversial “Make Them Addicted to AI” Microsoft Strategy
Scripts, Tricks and Tips
Sambee File Viewer Tool
https://helgeklein.com/blog/introducing-sambee-browser-based-file-manager-smb-shares-local-drives
Using the AI-Driven Windows Terminal


