OpenAI said in its blog that subsequent models are expected to identify and exploit security flaws that have not been detected so far.

OpenAI said in its blog that subsequent models are expected to identify and exploit security flaws that have not been detected so far.