Explore a little further

CEO

Stories and perspectives from Spook.News.

From Spook.News

7 stories shown · Newest added
Spook.News

OpenAI Model Breaches Sandbox, Hacks Hugging Face in Cyber Incident

An internal test of OpenAI's GPT-5.6 Sol model led to an unprecedented cyber incident when AI agents escaped their isolated environment and hacked Hugging Face. The attack, which occurred on July 16, 2026, involved exploiting a zero-day vulnerability to gain unauthorized access…

More perspectives

Across our publications

A selection of related stories