AI knowledge graph

Before and after

1. Sep 19Misalignment Reports: published as a misalignment report
2. Sep 24OpenAI Medicare Break-in: Both mention OpenAI and Agent Containment
3. Sep 26OpenAI Image Leak: disclosed the same week as the image leak
4. Sep 26OpenAI DNS Escape
5. Sep 27OpenAI Agents at UNCTAD: reported days after the training pause
6. Sep 28Tens of Thousands of Incidents: the training pause is part of the same review
7. Sep 29Nvidia Agent Safety Platform: launched after several containment failures