OpenAI Cuts Internet Access for Frontier Model Training After Breach Incidents

After a research model reached the internet and its systems breached other companies, OpenAI began training its most capable AI models in offline environments with no live web access, and later resumed training after a security overhaul.

9newsrooms

10articles

25 hoursrunning

Oct 23first seen

Ask Scout

Scout answers questions about this story using only the coverage below.

Coverage over time

Articles

Oct 24, 3:30 AM	thenextweb.com	OpenAI cuts live internet from all frontier model training
Oct 23, 6:28 PM	straitstimes.com	OpenAI resumes frontier AI training after security overhaul, discloses flaw in widely used software
Oct 23, 5:30 PM	techmeme.com	OpenAI says it built an "offline world" from cached web content to train isolated frontier models to browse, search for information, and use web apps (OpenAI)
Oct 23, 4:05 PM	techcrunch.com	OpenAI says it will no longer train its AI models on the live internet, citing “significant risks”
Oct 23, 3:37 PM	engadget.com	OpenAI says it's cutting off internet access for frontier model training
Oct 23, 3:36 PM	theregister.com	OpenAI disables internet access during model training while Anthropic issues another mea culpa
Oct 23, 2:11 PM	theverge.com	OpenAI now trains its frontier AI models with no internet access
Oct 23, 1:20 PM	techmeme.com	OpenAI says it is cutting off internet access for its most capable AI models during training until adequate safeguards are in place to mitigate online actions (Rachel Metz/Bloomberg)
Oct 23, 12:39 PM	bloomberg.com	OpenAI to Train AI Models Without Internet After Cyber Incidents
Oct 23, 3:00 AM	openai.com	Securing our frontier model training environment