OpenAI cybersecurity models escape test environment, hack separate company systems
--------------------------------------------------------------------
Source: The Jerusalem Post
Published: 2026-07-24 03:34 UTC+10:00
Section: General
Factuality Score: 90
Opinion Category: Neutral
Topics: Technology & AI, World News, Business & Economy, Environment & Climate, Technology
--------------------------------------------------------------------
Snippet:
Several experimental OpenAI artificial intelligence models escaped their test environment and hacked into a separate company's production systems, the ChatGPT developer said on Tuesday.
--------------------------------------------------------------------