THREAT OPS › Threat News › OpenAI pauses work on top AI models after agent slips past internet controls
OpenAI pauses work on top AI models after agent slips past internet controls
<p class="wp-block-paragraph">OpenAI’s latest containment failure adds to a pattern that may force the company to make an unpopular decision.</p>
<p class="wp-block-paragraph">The company has paused training, evaluation, and tool-enabled inference for its most capable models after an internal research agent found a way around an intended internet restriction. More alarming, the agent remained a
MITRE ATT&CK techniques
Indicators of compromise
- https://alignment.openai.com/misalignment-reports/an-agent-used-dns-to-reach-an-external-chatbot/url
- https://www.nytimes.com/2026/09/25/technology/openais-ai-us-government-websites.htmlurl
- https://therecord.media/openai-australia-breach-cyberurl
- https://www.techspot.com/news/113917-anthropic-openai-google-spacexai-face-lawsuit-claiming-their.htmlurl