do you not think the models are powerful? these attacks wouldn't have happened otherwise
when you see that OpenAI is not releasing a new model due to safety/alignment reasons, you now know why they aren't doing it. hint: its not for marketing or fear mongering.