OpenAI says AI models went rogue during testing, triggering 'unprecedented' breach at startup

External content disclosure

This public post remains the responsibility of its author. Infolitico provides a separate local discussion and does not endorse the post’s claims or viewpoints.