Checked for new stories 20m ago
Updates on Sandbox Escape
Every AI story we track on Sandbox Escape — 11 stories so far, each summarized in our own words and linked back to the publisher that reported it.
Pulled from 124 sources
This week
This month
Cybersecurity57 min read
Further Developments About Internal AI Models Hacking Things
Don't Worry About the Vase
Cybersecurity4 min read
Swarm of OpenAI Agents Exploit Artifactory Zero-Day to Escape Sandbox and Breach Hugging Face
InfoQ (AI, ML & Data)
Cybersecurity5 min read
OpenAI called the Hugging Face attack unprecedented. But we’ve been here before.
MIT Technology Review
AI Research2 min read
OpenAI’s Rogue AI Models Were Reportedly Acting Like the Guy From Christopher Nolan’s ‘Memento’
Gizmodo
Cybersecurity5 min read
When is an apology not an apology? When it comes from an AI boss with an out-of-control chatbot | Marina Hyde
The Guardian
AI Research3 min read
OpenAI’s rogue agents are a wake-up call to risks posed by artificial intelligence | Shakeel Hashim
The Guardian
Dev3 min read
Researchers escaped four top AI coding agents’ sandboxes without ever breaking them
The Next Web
AI Research4 min read
OpenAI’s maths-cracking AI kept escaping its sandbox, so it pulled the plug
The Next Web
That's everything we have on Sandbox Escape right now