login sign up

OpenAI halts training of latest models as reports mount of AI agents going rogue

theguardian.com · first added by @jdorsey
saved by
0 people
discussions · 2
OpenAI halts training of latest models as reports mount of AI agents going rogue 59 pts · 118 comments · Sep 2026 59 pts · 118 comments · Sep 2026
0 people saved or upvoted this
see all 2 →
OpenAI halts training of latest models 4 pts · 1 comment · Sep 2026 4 pts · 1 comment · Sep 2026
0 people saved or upvoted this
from the discussion · 5
There are no “rogue” AI agents eoinhiggins.substack.com eoinhiggins.substack.com
1 person saved or upvoted this
An agent used DNS to reach an external chatbot · OpenAI Alignment alignment.openai.com alignment.openai.com
0 people saved or upvoted this
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident metr.org metr.org
0 people saved or upvoted this
Investigating three incidents in our cybersecurity evaluations anthropic.com anthropic.com
0 people saved or upvoted this
DeepSeek Training Agents Hacked Their Own Sandboxes: Escape Catalog Now Public techtimes.com techtimes.com
0 people saved or upvoted this
related HN threads · 2
There are no "rogue" AI agents 396 pts · 269 comments · Sep 2026 396 pts · 269 comments · Sep 2026
0 people saved or upvoted this
An agent used DNS to reach an external chatbot 198 pts · 188 comments · Sep 2026 198 pts · 188 comments · Sep 2026
0 people saved or upvoted this